WEBVTT 00:00:00.080 --> 00:00:01.670 The following content is provided 00:00:01.670 --> 00:00:03.820 under a Creative Commons license. 00:00:03.820 --> 00:00:06.550 Your support will help MIT OpenCourseWare continue 00:00:06.550 --> 00:00:10.160 to offer high-quality educational resources for free. 00:00:10.160 --> 00:00:12.700 To make a donation or to view additional materials 00:00:12.700 --> 00:00:16.620 from hundreds of MIT courses, visit MIT OpenCourseWare 00:00:16.620 --> 00:00:17.275 at ocw.mit.edu. 00:00:26.006 --> 00:00:26.880 PROFESSOR: All right. 00:00:26.880 --> 00:00:28.013 Good afternoon, everyone. 00:00:30.590 --> 00:00:35.170 Today, we are going to have two topics. 00:00:35.170 --> 00:00:39.050 In the first half we are going to discuss visual illusions, 00:00:39.050 --> 00:00:41.240 and in the second half we are going 00:00:41.240 --> 00:00:46.550 to talk about prosthetic devices for the blind. 00:00:46.550 --> 00:00:49.050 What I would like to start with is, again, 00:00:49.050 --> 00:00:53.110 to send around the attendance sheet, 00:00:53.110 --> 00:00:56.180 so please put your name to it. 00:00:56.180 --> 00:00:58.360 And then, in addition, I'm going to send 00:00:58.360 --> 00:01:01.430 around some handouts, one for each person. 00:01:02.760 --> 00:01:04.769 I'm going to hand them out in both directions. 00:01:04.769 --> 00:01:05.269 Thank you. 00:01:08.810 --> 00:01:11.050 The handouts we're going to look at, 00:01:11.050 --> 00:01:14.250 I'll refer to them specifically as we go around. 00:01:14.250 --> 00:01:18.640 And I also will show pretty much the same handouts, figures 00:01:18.640 --> 00:01:21.040 in the handouts, on the screen. 00:01:22.060 --> 00:01:22.560 All right. 00:01:22.560 --> 00:01:26.820 So for the first half of the presentation today 00:01:26.820 --> 00:01:30.380 we are going to talk about visual illusions. 00:01:30.380 --> 00:01:35.430 Now, the main emphasis of what I'm going to talk about, 00:01:35.430 --> 00:01:39.690 it'll be an attempt to convey to you what 00:01:39.690 --> 00:01:44.220 kinds of procedures one can use to get a better sense of why 00:01:44.220 --> 00:01:46.935 we see in illusions what we see. 00:01:48.100 --> 00:01:51.220 Some of these approaches are strictly behavioral. 00:01:51.220 --> 00:01:57.380 And some of them have involved experiments carried out 00:01:57.380 --> 00:02:03.720 on a more thorough basis using electrophysiological procedures 00:02:03.720 --> 00:02:04.270 and so on. 00:02:04.270 --> 00:02:07.310 But those I'm not going to talk too much about. 00:02:07.310 --> 00:02:10.800 But I will relate what we can figure out 00:02:10.800 --> 00:02:17.380 through how the visual system works. 00:02:17.380 --> 00:02:22.330 Now, because of that, there are hundreds of illusions, 00:02:22.330 --> 00:02:25.440 and I will emphasize those that allow 00:02:25.440 --> 00:02:29.440 us to think about why we see what we see. 00:02:29.440 --> 00:02:32.450 Now, one of these illusory effects 00:02:32.450 --> 00:02:34.820 is the so-called Hermann grid illusion, 00:02:34.820 --> 00:02:36.650 and that's the one I'm going to start 00:02:36.650 --> 00:02:39.830 with because it's a fascinating illusion. 00:02:39.830 --> 00:02:42.880 If you look at this, what you see 00:02:42.880 --> 00:02:49.470 is that at the intersections you see sort of smudges. 00:02:49.470 --> 00:02:54.590 And so the big question arose as to why we see these smudges. 00:03:00.400 --> 00:03:03.480 Now, does everybody see those smudges? 00:03:03.480 --> 00:03:06.190 If you look at the handouts, I also 00:03:06.190 --> 00:03:13.700 show you-- pay attention only to the front page, 00:03:13.700 --> 00:03:18.340 the one on the left, which says the Hermann grid illusion. 00:03:18.340 --> 00:03:22.440 So if you look at that either on your handout 00:03:22.440 --> 00:03:24.870 or you look at it on the screen here, 00:03:24.870 --> 00:03:29.190 you see these smudges at the intersections, not where you're 00:03:29.190 --> 00:03:33.650 looking at, but in the smudges other than where 00:03:33.650 --> 00:03:34.725 you're fixating. 00:03:34.725 --> 00:03:36.100 Does everybody see those smudges? 00:03:38.030 --> 00:03:38.940 OK. 00:03:38.940 --> 00:03:42.780 So the question arose why we have these smudges. 00:03:42.780 --> 00:03:49.640 And back in the 1960s, an investigator 00:03:49.640 --> 00:03:52.380 came up with a rather clever idea 00:03:52.380 --> 00:03:54.580 of why we see these smudges. 00:03:54.580 --> 00:03:57.350 And the person who came up with this theory 00:03:57.350 --> 00:03:59.330 was a fellow called Baumgartner. 00:03:59.330 --> 00:04:03.380 And so what he proposed, that this illusory effect 00:04:03.380 --> 00:04:07.000 is not directly attributable to the center-surround 00:04:07.000 --> 00:04:11.010 organization or the receptive fields or retinal ganglion 00:04:11.010 --> 00:04:11.780 cells. 00:04:11.780 --> 00:04:17.579 So what he proposed is that when a retinal ganglion cell falls 00:04:17.579 --> 00:04:21.709 in the center like this, there's more inhibition in the surround 00:04:21.709 --> 00:04:25.720 than when it falls along not the intersections, 00:04:25.720 --> 00:04:27.350 but along the straight lines. 00:04:28.400 --> 00:04:31.540 And because of that, there's more inhibition here than here, 00:04:31.540 --> 00:04:33.770 and that's why we see the smudges. 00:04:33.770 --> 00:04:36.630 Now, that was a very attractive theory. 00:04:36.630 --> 00:04:41.940 And in fact, it has appeared and still exists in many textbooks 00:04:41.940 --> 00:04:46.610 that you read about illusions and about vision in general. 00:04:46.610 --> 00:04:50.460 Now, OK, so here is the exact description 00:04:50.460 --> 00:04:52.930 that I just made to tell you this. 00:04:52.930 --> 00:04:56.000 And of course, you would expect that if this is the case, 00:04:56.000 --> 00:05:00.240 you should also get the smudges if you'd reverse the contrast. 00:05:00.240 --> 00:05:02.390 And in fact, that is true. 00:05:02.390 --> 00:05:05.420 So here, once again, you see the smudges, maybe not quite 00:05:05.420 --> 00:05:06.310 as sharp. 00:05:06.310 --> 00:05:09.850 But in this case, it's whiter smudges, 00:05:09.850 --> 00:05:14.080 whereas in the previous case, this one, 00:05:14.080 --> 00:05:15.790 you have darker smudges. 00:05:15.790 --> 00:05:19.460 So that kind of fits with that basic idea. 00:05:19.460 --> 00:05:23.840 But then, subsequently, people began 00:05:23.840 --> 00:05:27.930 to play around with various conditions to see when you do 00:05:27.930 --> 00:05:32.100 and when you don't get the smudge effect 00:05:32.100 --> 00:05:35.960 and to see whether that would fit with this theory. 00:05:35.960 --> 00:05:38.810 Now, that's the basic tenet in just about everything 00:05:38.810 --> 00:05:42.230 that psychologists and neuroscientists 00:05:42.230 --> 00:05:44.050 working in the visual system do. 00:05:44.050 --> 00:05:48.480 They try to critically examine various hypotheses. 00:05:49.610 --> 00:05:51.200 And doesn't have to be just illusions, 00:05:51.200 --> 00:05:53.370 it can be a whole bunch of other things. 00:05:53.370 --> 00:05:56.940 And they do this then by systematically manipulating 00:05:56.940 --> 00:06:02.050 things in experiments that can present these things just 00:06:02.050 --> 00:06:05.580 under free viewing conditions, as I'm doing it 00:06:05.580 --> 00:06:08.250 in terms of these handouts, or you 00:06:08.250 --> 00:06:10.640 can do it under more controlled conditions. 00:06:10.640 --> 00:06:13.690 The subjects can look through a tachistoscope or something 00:06:13.690 --> 00:06:14.640 like that. 00:06:14.640 --> 00:06:15.230 All right. 00:06:15.230 --> 00:06:19.780 So anyway, let's now look at this in a bit more detail 00:06:19.780 --> 00:06:24.050 and see what's wrong with this theory. 00:06:24.050 --> 00:06:27.780 Well, first of all, if you look at this figure, 00:06:27.780 --> 00:06:30.370 what you can see here that you can see the smudges 00:06:30.370 --> 00:06:33.645 irrespective of the size of the overall display. 00:06:34.780 --> 00:06:37.290 Now, that's a problem because, of course, 00:06:37.290 --> 00:06:40.640 the size of the receptive fields in the retina, 00:06:40.640 --> 00:06:44.530 in their center-surround arrangement, is constant. 00:06:44.530 --> 00:06:47.700 So if it were the sharp arrangement 00:06:47.700 --> 00:06:51.510 that we described about the way the center and the surround 00:06:51.510 --> 00:06:53.810 are activated by the intersections 00:06:53.810 --> 00:06:57.880 and the non-intersections in the display, that should not 00:06:57.880 --> 00:06:59.460 apply to different sizes. 00:06:59.460 --> 00:07:03.260 It should only work then for a particular size 00:07:03.260 --> 00:07:05.400 where you approximate that. 00:07:05.400 --> 00:07:09.640 Now, I will elaborate on that in more detail in just a minute. 00:07:09.640 --> 00:07:12.440 Now, another bug that arose here is 00:07:12.440 --> 00:07:17.130 that if you now turn this display just 45 degrees, 00:07:17.130 --> 00:07:20.045 all right, so that it's a diamond rather than a square, 00:07:20.045 --> 00:07:23.765 the illusory effect has declined considerably. 00:07:24.960 --> 00:07:27.370 And then to make this even more problematic, 00:07:27.370 --> 00:07:32.750 and this appears in your handout on that backside 00:07:32.750 --> 00:07:37.110 of the first page, what you can see here 00:07:37.110 --> 00:07:40.980 is that when you make the lines arranged 00:07:40.980 --> 00:07:45.060 in such a fashion at slight angles to each other, then 00:07:45.060 --> 00:07:47.540 you don't see the illusory effect at all. 00:07:47.540 --> 00:07:56.690 Whereas here, we also see it a little bit less but still 00:07:56.690 --> 00:07:58.470 slightly visible. 00:07:58.470 --> 00:08:00.350 But here it's not visible at all. 00:08:00.350 --> 00:08:04.600 And you can compare that even better on the handout there. 00:08:04.600 --> 00:08:07.520 So the question is, why would that be? 00:08:08.880 --> 00:08:10.580 Well, we give you another example here. 00:08:10.580 --> 00:08:12.260 If you make serrated edges-- I don't 00:08:12.260 --> 00:08:15.090 know if the people in the back can see the serrations. 00:08:15.090 --> 00:08:17.430 Those of you in front can see the serrations. 00:08:17.430 --> 00:08:20.150 And when you do that, also you fail to see it. 00:08:20.150 --> 00:08:22.290 And the problem here, of course, is 00:08:22.290 --> 00:08:26.490 that if you make these slight changes in the angles 00:08:26.490 --> 00:08:32.909 of the intersecting lines or you make them serrated, 00:08:32.909 --> 00:08:36.860 the center-surround antagonism should be the same. 00:08:36.860 --> 00:08:41.289 And yet such a small change causes the illusory effect 00:08:41.289 --> 00:08:42.590 to disappear. 00:08:42.590 --> 00:08:46.830 So that then, in itself, questions the Baumgartner 00:08:46.830 --> 00:08:47.330 theory. 00:08:49.050 --> 00:08:52.040 Now, there are several other factors that enter into this. 00:08:52.040 --> 00:08:53.850 This is a very subtle one. 00:08:53.850 --> 00:08:57.710 If you look at this side, OK, we have different shades 00:08:57.710 --> 00:08:58.991 for the lines. 00:08:58.991 --> 00:08:59.490 OK? 00:09:01.180 --> 00:09:05.610 On this side, we have the darkened lines, OK, in front. 00:09:05.610 --> 00:09:07.760 And in this case, they have them in the back. 00:09:07.760 --> 00:09:10.930 And here you see the illusory effect, and here you don't. 00:09:10.930 --> 00:09:14.330 And once again, how can that be if the illusory effect is 00:09:14.330 --> 00:09:16.770 due to the center-surround antagonism? 00:09:16.770 --> 00:09:20.600 Even more dramatic is when you do this with colors. 00:09:20.600 --> 00:09:23.782 Here you have, again, the colored lines in front, 00:09:23.782 --> 00:09:25.240 and here you have them in the back. 00:09:25.240 --> 00:09:30.500 And here you can see homonymous shades, all right? 00:09:30.500 --> 00:09:33.760 This is darker yellow and so on. 00:09:34.800 --> 00:09:37.860 But it's the same color but a little bit darker 00:09:37.860 --> 00:09:39.310 at the intersections. 00:09:39.310 --> 00:09:41.580 So here you get a pretty good illusory effect. 00:09:41.580 --> 00:09:44.800 But once those white lines are in front, 00:09:44.800 --> 00:09:47.960 we get very little effect at all. 00:09:47.960 --> 00:09:53.020 So that then denotes a number of problems with the theory. 00:09:53.020 --> 00:09:57.340 And then one more fact is that when you present the stimuli 00:09:57.340 --> 00:10:01.210 at isoluminance, you don't get an effect at all. 00:10:01.210 --> 00:10:03.940 So that then raised some serious questions 00:10:03.940 --> 00:10:05.720 about the illusory effect. 00:10:05.720 --> 00:10:08.390 And then a quantitative study was 00:10:08.390 --> 00:10:13.150 undertaken to see just what is the layout of the receptive 00:10:13.150 --> 00:10:16.790 fields in this display. 00:10:16.790 --> 00:10:20.090 And so here what we have is a display. 00:10:20.090 --> 00:10:25.830 This square here is exactly five degrees from your fixation, OK, 00:10:25.830 --> 00:10:31.610 and you can calculate the number of midget and parasol cells 00:10:31.610 --> 00:10:35.250 that are activated by that square area. 00:10:35.250 --> 00:10:39.570 And you can see in total there are more than 360 cells that 00:10:39.570 --> 00:10:42.390 are being activated in that little region. 00:10:42.390 --> 00:10:48.540 And here is a display, in this case, of only the ON cells, 00:10:48.540 --> 00:10:50.760 and you can double that by having the OFF 00:10:50.760 --> 00:10:52.570 cells in the same location. 00:10:52.570 --> 00:10:57.100 Saying that there's just no good correspondence 00:10:57.100 --> 00:11:02.210 whatsoever in the number of ganglion cells 00:11:02.210 --> 00:11:05.220 that are being activated and how the center-surround is truly 00:11:05.220 --> 00:11:11.230 laid out at a point where you get a very good smudge 00:11:11.230 --> 00:11:15.030 effect with this illusion. 00:11:15.030 --> 00:11:17.850 So this then, taken all together, 00:11:17.850 --> 00:11:27.380 raises the issue then if indeed there's a retina phenomenon, 00:11:27.380 --> 00:11:29.240 why is there no fit here? 00:11:29.240 --> 00:11:31.660 And so it does look like that it's unlikely 00:11:31.660 --> 00:11:33.760 that this illusion can be explained 00:11:33.760 --> 00:11:36.070 by the center-surround antagonism theory. 00:11:39.780 --> 00:11:42.460 So then another theory was advanced 00:11:42.460 --> 00:11:45.150 claiming that this doesn't take place in the retina, 00:11:45.150 --> 00:11:47.930 but instead that it takes place in the cortex. 00:11:49.200 --> 00:11:51.560 And so another theory that was advanced 00:11:51.560 --> 00:11:54.780 is saying that this happens in the cortex due to the fact 00:11:54.780 --> 00:11:58.090 that you have these simple cells-- we talked 00:11:58.090 --> 00:12:05.530 about those before-- that had these elongated inhibitory, 00:12:05.530 --> 00:12:10.500 supposedly inhibitory, surrounds. 00:12:10.500 --> 00:12:15.600 And if one calculates this, that can explain the illusion 00:12:15.600 --> 00:12:19.390 a little bit better, especially since these cells, 00:12:19.390 --> 00:12:22.310 even at any given point in the visual cortex, 00:12:22.310 --> 00:12:23.755 come in several different sizes. 00:12:24.940 --> 00:12:26.970 So that's an alternate theory. 00:12:26.970 --> 00:12:28.780 But still, further research needs 00:12:28.780 --> 00:12:34.980 to be done to be able to verify whether this theory is more 00:12:34.980 --> 00:12:41.040 valid than the center-surround theory in the retina, which 00:12:41.040 --> 00:12:43.390 certainly is totally wrong. 00:12:43.390 --> 00:12:47.200 So I think maybe in another 100 years, 00:12:47.200 --> 00:12:50.280 they're going to drop the Baumgartner theory out 00:12:50.280 --> 00:12:58.190 from textbooks and general reviews of illusions 00:12:58.190 --> 00:12:59.150 and vision. 00:12:59.150 --> 00:12:59.650 OK. 00:12:59.650 --> 00:13:03.560 So now I want to show you one more illusory effect, which 00:13:03.560 --> 00:13:13.370 is a variant of the illusion that I have just shown you. 00:13:13.370 --> 00:13:17.180 And this one is something that can drive you nuts, I suppose. 00:13:17.180 --> 00:13:23.970 And that appears on the second front page, the second sheet, 00:13:23.970 --> 00:13:26.330 I should say, of the handout. 00:13:26.330 --> 00:13:29.100 You can look at it there, or you can look at it here. 00:13:29.100 --> 00:13:34.020 And you can see this scintillating effect, 00:13:34.020 --> 00:13:38.380 and therefore it has been called the scintillating grid 00:13:38.380 --> 00:13:39.310 illusion. 00:13:39.310 --> 00:13:41.320 It was invented by fellow called Lingelbach. 00:13:42.340 --> 00:13:43.080 All right? 00:13:43.080 --> 00:13:46.450 That's a very strong illusory effect, very dramatic. 00:13:46.450 --> 00:13:49.360 And whether this can be explained 00:13:49.360 --> 00:13:54.070 by any particular theory about the visual cortex 00:13:54.070 --> 00:13:55.810 still remains to be settled. 00:13:55.810 --> 00:14:00.450 Right now I don't think there's a specific theory that 00:14:00.450 --> 00:14:03.730 can explain this. 00:14:03.730 --> 00:14:08.380 Now, I will now move on and talk about yet 00:14:08.380 --> 00:14:12.740 another illusory effect, or actually a set of them, 00:14:12.740 --> 00:14:15.960 that has a much better explanation. 00:14:15.960 --> 00:14:18.080 This one we don't have a handout for. 00:14:18.080 --> 00:14:20.630 If you look at this display here, 00:14:20.630 --> 00:14:24.560 you'll agree that the left and right bars here are identical, 00:14:24.560 --> 00:14:27.740 right, going in shade from light to dark. 00:14:27.740 --> 00:14:28.330 All right? 00:14:28.330 --> 00:14:32.520 So now what I'm going to do here-- what I want to do 00:14:32.520 --> 00:14:35.170 is to present these two images here. 00:14:35.170 --> 00:14:37.730 I want you to fixate in the fixation spot, 00:14:37.730 --> 00:14:41.230 and we'll count to about, I don't know, 20, say. 00:14:41.230 --> 00:14:42.570 Keep fixating. 00:14:42.570 --> 00:14:45.300 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. 00:14:45.300 --> 00:14:47.640 1, 2, 3, 4, 5, 6, 7, 8, 9, 20. 00:14:47.640 --> 00:14:49.390 And then I'm going to go back here, 00:14:49.390 --> 00:14:51.890 and you can see that this side is much darker than that one. 00:14:51.890 --> 00:14:53.105 Everybody got that effect? 00:14:53.105 --> 00:14:53.580 AUDIENCE: Yeah. 00:14:53.580 --> 00:14:54.913 PROFESSOR: I'll do it once more. 00:14:59.050 --> 00:15:00.401 Pretty dramatic, huh? 00:15:00.401 --> 00:15:00.900 OK. 00:15:00.900 --> 00:15:02.524 Now I'll tell you what I'm going to do. 00:15:04.140 --> 00:15:08.670 What I want you to do is to look at this with one eye. 00:15:08.670 --> 00:15:09.290 OK? 00:15:09.290 --> 00:15:10.500 Don't look at it yet. 00:15:10.500 --> 00:15:11.620 Wait a while. 00:15:11.620 --> 00:15:12.950 Just let your eye get settled. 00:15:14.180 --> 00:15:16.000 You're going to look at it with one eye. 00:15:16.000 --> 00:15:19.500 And then when I switch over, I want you to switch eyes. 00:15:19.500 --> 00:15:20.000 OK? 00:15:21.062 --> 00:15:21.770 All right, ready? 00:15:21.770 --> 00:15:25.520 Go ahead now and fixate, and we'll count to 20 again. 00:15:25.520 --> 00:15:27.020 5, 6, 7, 8, 9, 10. 00:15:27.020 --> 00:15:29.655 1, 2, 3, 4, 5, 6, 7, 8, 9, 20. 00:15:29.655 --> 00:15:30.155 Switch. 00:15:31.860 --> 00:15:34.500 And if you do that, you don't have an effect. 00:15:34.500 --> 00:15:35.000 OK? 00:15:35.000 --> 00:15:36.760 So what does that prove? 00:15:36.760 --> 00:15:38.846 Well, that obviously proves-- that's why 00:15:38.846 --> 00:15:41.780 it's nice to do these little, itty-bitty experiments one can 00:15:41.780 --> 00:15:46.490 do in minutes-- it proves that this is a retinal effect. 00:15:46.490 --> 00:15:50.020 And the effect is actually due to adaptation 00:15:50.020 --> 00:15:55.020 that we have talked about, when was it, a couple of times ago. 00:15:55.020 --> 00:15:58.140 Now, I want to show you yet another effect 00:15:58.140 --> 00:16:01.440 that has a similar arrangement. 00:16:01.440 --> 00:16:06.920 What I'm going to do here is I'm going to put this display on. 00:16:06.920 --> 00:16:09.080 And what I'm going to do here is I'm 00:16:09.080 --> 00:16:11.040 going to turn these off successively 00:16:11.040 --> 00:16:12.780 as we go around the clock. 00:16:12.780 --> 00:16:13.440 OK? 00:16:13.440 --> 00:16:15.420 And what I want you to do, again, 00:16:15.420 --> 00:16:18.660 is to fixate here on the cross, OK, 00:16:18.660 --> 00:16:21.490 and then I'm going to start running it. 00:16:21.490 --> 00:16:23.740 And what are you going to see? 00:16:23.740 --> 00:16:26.334 A pretty amazing thing you're going to see. 00:16:26.334 --> 00:16:27.000 What do you see? 00:16:27.000 --> 00:16:27.916 AUDIENCE: [INAUDIBLE]. 00:16:29.410 --> 00:16:30.740 PROFESSOR: Yeah? 00:16:30.740 --> 00:16:33.080 You see a green spot going around? 00:16:33.080 --> 00:16:35.061 Everybody see that? 00:16:35.061 --> 00:16:35.560 OK. 00:16:35.560 --> 00:16:39.650 So now the big question comes up, why on earth 00:16:39.650 --> 00:16:41.715 do we see this green spot? 00:16:41.715 --> 00:16:42.715 What is the explanation? 00:16:44.160 --> 00:16:47.240 Well, this again is a retinal effect. 00:16:47.240 --> 00:16:53.410 And to test this in more detail, we 00:16:53.410 --> 00:16:57.780 have to go back to what I told you about color vision. 00:16:57.780 --> 00:16:59.960 And when I told you about color vision-- 00:16:59.960 --> 00:17:04.349 if you remember we did that experiment with Gaussians-- 00:17:04.349 --> 00:17:10.220 that if you look at the color circle, all right, and then 00:17:10.220 --> 00:17:15.710 you adapt to a stimulus here, which is pretty similar to what 00:17:15.710 --> 00:17:18.810 was on the screen, then the afterimage 00:17:18.810 --> 00:17:24.740 is on the opposite side cutting across the color circle. 00:17:24.740 --> 00:17:30.490 So the afterimages that you see that are in color 00:17:30.490 --> 00:17:33.770 can be perfectly explained by the color circle. 00:17:33.770 --> 00:17:35.830 If you adapt to this, this is the afterimage. 00:17:35.830 --> 00:17:38.420 You adapt to this, this is the afterimage. 00:17:38.420 --> 00:17:44.900 It's always straight across the center line, center point, 00:17:44.900 --> 00:17:49.780 I should say, where you have the white center, which perfectly 00:17:49.780 --> 00:17:53.620 explains your afterimage. 00:17:53.620 --> 00:17:57.075 So what I'll do again, I'll show this to you just once more. 00:18:00.750 --> 00:18:02.720 OK, here we go, ready? 00:18:02.720 --> 00:18:07.360 You can experience it again just to sort of get a sense. 00:18:07.360 --> 00:18:09.690 It's amazing how the green circle 00:18:09.690 --> 00:18:11.639 becomes more and more and more and more vivid. 00:18:11.639 --> 00:18:13.180 Now, one more thing I'm going to tell 00:18:13.180 --> 00:18:16.020 you is when I turn this off-- keep fixating-- 00:18:16.020 --> 00:18:18.960 and then you're going to see a green afterimage, right? 00:18:18.960 --> 00:18:23.270 So it proves that, indeed, it's the afterimage that 00:18:23.270 --> 00:18:26.860 appears over a period of time with that rotating circle, 00:18:26.860 --> 00:18:30.195 and it's fully explained by the color circle. 00:18:31.701 --> 00:18:32.200 All right. 00:18:32.200 --> 00:18:35.350 So now we're going to move on, and you 00:18:35.350 --> 00:18:37.500 look at yet another illusory effect. 00:18:37.500 --> 00:18:41.080 If you look at this display, what do you see here? 00:18:41.080 --> 00:18:44.980 You see a light square and a dark square, right? 00:18:44.980 --> 00:18:45.990 Everybody see that? 00:18:47.350 --> 00:18:51.490 Now, what if I told you that what you're seeing 00:18:51.490 --> 00:18:53.696 is a total misconception? 00:18:54.567 --> 00:18:56.150 So I'll tell you what I'm going to do. 00:18:56.150 --> 00:18:59.160 I'm going to leave these two squares exactly as they are, 00:18:59.160 --> 00:19:00.860 but I'm going to change the background. 00:19:00.860 --> 00:19:01.360 Ready? 00:19:03.386 --> 00:19:04.510 What's the difference here? 00:19:05.580 --> 00:19:10.110 The difference is that now the background is uniform, right? 00:19:10.110 --> 00:19:12.270 And you can see these two squares are indeed 00:19:12.270 --> 00:19:14.530 identical in contrast, right? 00:19:14.530 --> 00:19:17.081 But now if I go back to the previous one, 00:19:17.081 --> 00:19:18.580 if you look at it more carefully you 00:19:18.580 --> 00:19:20.230 can see this goes from light to dark. 00:19:21.290 --> 00:19:23.650 Of course, the projector makes this look bluish. 00:19:23.650 --> 00:19:25.870 It shouldn't be, but that's OK. 00:19:25.870 --> 00:19:30.470 And it's a gradual shift in the illumination level 00:19:30.470 --> 00:19:35.320 that fools you to think that one is light, the other is dark. 00:19:35.320 --> 00:19:38.050 And that has to do heavily with the fact 00:19:38.050 --> 00:19:41.780 that the receptive fields of cells in the retina 00:19:41.780 --> 00:19:44.210 have center-surround organization, 00:19:44.210 --> 00:19:47.740 and they're sensitive to local contrast differences. 00:19:47.740 --> 00:19:48.880 That's their make-up. 00:19:48.880 --> 00:19:51.870 And so until I point it out to you, 00:19:51.870 --> 00:19:55.960 can't even tell that this is a gradual shift. 00:19:55.960 --> 00:19:58.540 And if it did not change in color, 00:19:58.540 --> 00:20:00.850 it would be harder to tell. 00:20:00.850 --> 00:20:02.600 And it has to do with the fact that you're 00:20:02.600 --> 00:20:04.900 making local comparisons. 00:20:04.900 --> 00:20:09.240 And because of that, to make a distant comparison 00:20:09.240 --> 00:20:11.300 you are inaccurate about it. 00:20:11.300 --> 00:20:14.680 So this illusory effect, again, can be readily explained 00:20:14.680 --> 00:20:18.410 on the basis of what we know about the center-surround 00:20:18.410 --> 00:20:21.951 organization of receptive fields in the retina. 00:20:21.951 --> 00:20:22.450 OK. 00:20:22.450 --> 00:20:26.780 Now, if you look at the same thing in color, which 00:20:26.780 --> 00:20:30.450 doesn't come up too well here because of the projector, 00:20:30.450 --> 00:20:34.270 but still you can tell that these are not the same color, 00:20:34.270 --> 00:20:37.160 OK, again because of the gradual shift, 00:20:37.160 --> 00:20:38.830 color shift, in the background. 00:20:38.830 --> 00:20:43.190 But exactly identical they are, as they had been before. 00:20:43.190 --> 00:20:46.660 So this works not only for contrast, 00:20:46.660 --> 00:20:48.750 but it also works for color. 00:20:48.750 --> 00:20:53.441 Now, some illusory effects don't work for color and some do. 00:20:53.441 --> 00:20:53.940 All right. 00:20:53.940 --> 00:20:58.460 So now yet another related effect 00:20:58.460 --> 00:21:00.290 that I'm sure some of you have seen 00:21:00.290 --> 00:21:03.870 before is if you look at this circle, 00:21:03.870 --> 00:21:06.310 we agree that it's uniform in color. 00:21:06.310 --> 00:21:11.119 Now, if I cover half of it with dark and the other with light, 00:21:11.119 --> 00:21:12.410 it still looks pretty the same. 00:21:12.410 --> 00:21:14.410 Maybe it's a little darker here than here. 00:21:14.410 --> 00:21:18.880 But then if I put a partition here, as you can see here, 00:21:18.880 --> 00:21:22.750 then this side definitely looks darker than that side. 00:21:22.750 --> 00:21:23.820 OK? 00:21:23.820 --> 00:21:26.430 So this, again, looks at contrast effects, 00:21:26.430 --> 00:21:30.577 and we'll talk about these contrasts in just a bit 00:21:30.577 --> 00:21:31.285 with more detail. 00:21:32.420 --> 00:21:36.060 Next, let me turn to another illusory effect 00:21:36.060 --> 00:21:38.840 that you can see in many, many textbooks, which 00:21:38.840 --> 00:21:40.960 is called the wallpaper illusion. 00:21:40.960 --> 00:21:44.500 If you look at this-- I'll come to it 00:21:44.500 --> 00:21:49.780 in a second in your display, where it will appear 00:21:49.780 --> 00:21:55.250 on the back of the second page-- what you see here, 00:21:55.250 --> 00:21:59.770 it looks like that these are not parallel lines here. 00:21:59.770 --> 00:22:03.300 I shouldn't say lines, but parallel 00:22:03.300 --> 00:22:05.480 black and white squares. 00:22:05.480 --> 00:22:08.810 It looks like that they're going like back and forth like that, 00:22:08.810 --> 00:22:10.640 OK, and it's quite dramatic. 00:22:10.640 --> 00:22:13.710 And so now we can ask the question, how can we see this? 00:22:13.710 --> 00:22:15.650 What is this due to? 00:22:15.650 --> 00:22:18.582 It could be due to the way they're aligned. 00:22:18.582 --> 00:22:20.165 It could be due to the fact that there 00:22:20.165 --> 00:22:22.670 is space between them, that gray space. 00:22:22.670 --> 00:22:26.570 And so one can do experiments like the one 00:22:26.570 --> 00:22:29.930 you have on your next page there. 00:22:29.930 --> 00:22:32.010 Let me first-- I will prove that, indeed, we 00:22:32.010 --> 00:22:33.250 have parallel lines here. 00:22:33.250 --> 00:22:36.080 These are fully parallel lines, and they go back and forth. 00:22:36.080 --> 00:22:39.900 To convince you, you can see that the lines are identical, 00:22:39.900 --> 00:22:43.110 and yet they do not look parallel 00:22:43.110 --> 00:22:45.881 at all because of this illusory effect. 00:22:45.881 --> 00:22:46.380 OK. 00:22:46.380 --> 00:22:50.882 So now what we can do is we can fiddle around with this a bit. 00:22:50.882 --> 00:22:53.090 Now, that's the kind of thing you guys would probably 00:22:53.090 --> 00:22:55.360 enjoy doing yourselves-- look at various illusions 00:22:55.360 --> 00:22:57.110 and start playing around with them. 00:22:57.110 --> 00:22:58.120 This is an example. 00:22:58.120 --> 00:23:00.620 This is a basic wallpaper illusion. 00:23:00.620 --> 00:23:03.640 And now you turn this into an equal checkerboard, OK? 00:23:03.640 --> 00:23:07.280 And if you keep the lines in between, 00:23:07.280 --> 00:23:08.720 and you get no effect whatsoever. 00:23:09.830 --> 00:23:13.080 And then here we eliminate those gray bars, 00:23:13.080 --> 00:23:15.760 and the illusory effect is greatly reduced. 00:23:15.760 --> 00:23:18.610 So consequently, the illusory effect 00:23:18.610 --> 00:23:21.150 has to do with both the fact that you 00:23:21.150 --> 00:23:25.470 have space between the alternating black and white 00:23:25.470 --> 00:23:29.110 bars, meaning these gray lines here, 00:23:29.110 --> 00:23:34.390 and that you have them offset. 00:23:34.390 --> 00:23:36.605 So when it's not offset, you don't get it. 00:23:36.605 --> 00:23:38.730 And when you don't have the bars, you don't get it. 00:23:38.730 --> 00:23:42.590 So these two factors play an important role 00:23:42.590 --> 00:23:46.340 in giving rise to the illusory effect. 00:23:46.340 --> 00:23:51.370 Now, what we also can examine is what happens at isoluminance. 00:23:52.590 --> 00:23:55.030 And in this case, once again, what happens 00:23:55.030 --> 00:23:57.470 is that the illusory effect is still there, 00:23:57.470 --> 00:24:00.960 but it is reduced under isoluminant conditions. 00:24:00.960 --> 00:24:03.230 And this is not the isoluminant actually 00:24:03.230 --> 00:24:04.730 because of the projector. 00:24:04.730 --> 00:24:06.640 But when it's perfectly isoluminant, 00:24:06.640 --> 00:24:09.500 then the illusory effect is largely lost. 00:24:10.780 --> 00:24:11.280 All right. 00:24:11.280 --> 00:24:15.510 So now we have a variant of this illusion which is fascinating. 00:24:15.510 --> 00:24:22.300 And this one I will show you on the next page, OK? 00:24:22.300 --> 00:24:25.350 And if you look at just this display, 00:24:25.350 --> 00:24:29.200 and you are convinced that we have lines that are alternating 00:24:29.200 --> 00:24:32.000 that are filtered back and forth, right? 00:24:33.770 --> 00:24:35.760 Well, is that really true? 00:24:35.760 --> 00:24:41.320 Well, it turns out if you take a bunch of red lines here 00:24:41.320 --> 00:24:43.820 that are perfectly parallel, OK, and then you 00:24:43.820 --> 00:24:47.720 superimpose those on the display, 00:24:47.720 --> 00:24:51.220 you can see that actually the display is parallel. 00:24:51.220 --> 00:24:52.210 OK? 00:24:52.210 --> 00:24:54.650 The reason it doesn't look parallel 00:24:54.650 --> 00:24:59.795 is due to the way these individual little pieces 00:24:59.795 --> 00:25:03.270 that you're not even conscious of are arranged. 00:25:03.270 --> 00:25:04.945 So let's look at that in more detail. 00:25:06.520 --> 00:25:08.400 Let me add here one more thing. 00:25:08.400 --> 00:25:11.300 At isoluminance, you get practically no effect. 00:25:11.300 --> 00:25:13.790 So these illusions, this one and the previous one, 00:25:13.790 --> 00:25:17.200 heavily depend on contrast differences. 00:25:17.200 --> 00:25:17.700 OK. 00:25:17.700 --> 00:25:18.780 So here we go. 00:25:18.780 --> 00:25:20.710 Here's the basic illusory effect. 00:25:20.710 --> 00:25:24.820 And the way this was constructed is shown enlarged on top. 00:25:24.820 --> 00:25:31.460 Alternate lines have 1, 2, 1, 2, 1, 2, 1. 00:25:31.460 --> 00:25:32.060 OK? 00:25:32.060 --> 00:25:36.810 What you have in the top line, and the third, and so on, 00:25:36.810 --> 00:25:38.320 you have it in this orientation. 00:25:38.320 --> 00:25:40.220 Here you have them reversed. 00:25:40.220 --> 00:25:44.560 And that vague cue is what somehow 00:25:44.560 --> 00:25:47.080 provides the illusory effect for you. 00:25:47.080 --> 00:25:51.090 Now, if you make all the lines similar, 00:25:51.090 --> 00:25:54.610 OK, in other words don't reverse the alternate lines, 00:25:54.610 --> 00:25:58.750 then what you see here is actually 00:25:58.750 --> 00:26:01.190 the bars are all tilted slightly upward. 00:26:01.190 --> 00:26:01.970 OK? 00:26:01.970 --> 00:26:03.690 And because of that, the illusory effect 00:26:03.690 --> 00:26:06.830 is not as dramatic as this one, but you still have it. 00:26:06.830 --> 00:26:07.590 All right? 00:26:07.590 --> 00:26:14.090 So what then you can do with this 00:26:14.090 --> 00:26:19.090 is to enlarge on this effect as shown here. 00:26:19.090 --> 00:26:23.515 It looks like that you have somehow a very unequal set 00:26:23.515 --> 00:26:30.860 of lines, as a result of which you do not 00:26:30.860 --> 00:26:33.140 have a bunch of clear-cut squares. 00:26:33.140 --> 00:26:37.380 And if you, again, superimpose true parallel lines through it, 00:26:37.380 --> 00:26:40.010 you see that they are actually parallel. 00:26:40.010 --> 00:26:40.510 All right. 00:26:40.510 --> 00:26:46.060 So that then explains that illusory effect 00:26:46.060 --> 00:26:51.760 in terms of how we somehow integrate information of shape 00:26:51.760 --> 00:26:55.790 that we are not even aware of to have 00:26:55.790 --> 00:26:57.985 a perception of a whole thing. 00:26:57.985 --> 00:26:59.610 Now, that's not much of an explanation. 00:26:59.610 --> 00:27:02.194 It doesn't tell you what neurons are involved 00:27:02.194 --> 00:27:03.110 or anything like that. 00:27:03.110 --> 00:27:05.350 But it gives you a general concept 00:27:05.350 --> 00:27:07.730 of how that illusory effect is created. 00:27:12.020 --> 00:27:17.800 Now we come to yet another fascinating illusory effect, 00:27:17.800 --> 00:27:23.660 and that one is in the next few pages in the handout. 00:27:23.660 --> 00:27:28.250 So if you take the backside of page 3, 00:27:28.250 --> 00:27:32.010 you will see this illusory effect here. 00:27:32.010 --> 00:27:33.470 Does everybody see that? 00:27:33.470 --> 00:27:38.100 Now, this illusory effect is one which gives you a sense 00:27:38.100 --> 00:27:39.950 that you have a whole bunch of swirls. 00:27:39.950 --> 00:27:40.942 All right? 00:27:40.942 --> 00:27:42.275 Does everybody see those swirls? 00:27:44.310 --> 00:27:49.780 Instead of seeing 1, 2, 3, 4 circles, you see some swirls. 00:27:49.780 --> 00:27:54.530 And you say what on earth is giving us that sensation? 00:27:54.530 --> 00:27:58.670 So what you want to do then is to look at it analytically. 00:27:58.670 --> 00:28:00.780 And what you see here is you have 00:28:00.780 --> 00:28:04.870 alternating black and white squares, first of all. 00:28:04.870 --> 00:28:06.880 And you say, well, is alternating 00:28:06.880 --> 00:28:10.170 black and white squares the central point? 00:28:10.170 --> 00:28:12.550 But then if you look at it more closely, 00:28:12.550 --> 00:28:14.810 you can see this is-- I don't know how they came up 00:28:14.810 --> 00:28:17.950 with this one-- what you see here, if you look at each 00:28:17.950 --> 00:28:20.440 of these, that there's a gradual shift as you go 00:28:20.440 --> 00:28:25.440 around the circle in the orientation of the square. 00:28:25.440 --> 00:28:27.550 So this square becomes progressively, 00:28:27.550 --> 00:28:31.420 very slowly, more rotated. 00:28:31.420 --> 00:28:32.050 OK? 00:28:32.050 --> 00:28:33.799 And that goes around the clock, and that's 00:28:33.799 --> 00:28:35.715 true for every one of those circles. 00:28:35.715 --> 00:28:42.740 And you can say, well, is it due to the black and white squares? 00:28:42.740 --> 00:28:46.310 Is it due to the change in the orientation of the squares 00:28:46.310 --> 00:28:48.230 that you're not even aware of? 00:28:48.230 --> 00:28:49.050 Or what? 00:28:49.050 --> 00:28:50.880 So what kind of experiment do you 00:28:50.880 --> 00:28:54.430 do to check this out, to come up with some understanding? 00:28:54.430 --> 00:28:56.820 What kind of game can you play? 00:28:56.820 --> 00:28:58.970 Well, the first game you can play, 00:28:58.970 --> 00:29:00.840 so you can say, well, let's eliminate 00:29:00.840 --> 00:29:04.130 these slight changes in orientation. 00:29:05.280 --> 00:29:05.910 OK? 00:29:05.910 --> 00:29:10.150 And so what you can do instead, as shown on the next page, 00:29:10.150 --> 00:29:15.070 is that you have a bunch of circles, 00:29:15.070 --> 00:29:18.290 black and white circles, which are all the same. 00:29:18.290 --> 00:29:21.110 And as you can see here, there's no swirls. 00:29:21.110 --> 00:29:26.000 You have four circles, very clear. 00:29:26.000 --> 00:29:29.670 So therefore, obviously, the squares, 00:29:29.670 --> 00:29:32.670 changing their orientation in a subtle fashion going around 00:29:32.670 --> 00:29:36.120 in the clock, is very important in introducing 00:29:36.120 --> 00:29:37.600 this illusory effect. 00:29:37.600 --> 00:29:42.050 Well, then you can say, OK, but is that all of it? 00:29:42.050 --> 00:29:45.130 And so what we can do here, as shown on the next page, 00:29:45.130 --> 00:29:48.300 is we can maintain the orientation shift, 00:29:48.300 --> 00:29:55.820 but we can make them all the same contrast, as we have here. 00:29:55.820 --> 00:29:58.700 And when you do that, you have have a rather weak illusory 00:29:58.700 --> 00:29:59.200 effect. 00:29:59.200 --> 00:30:02.720 There's still a little bit left maybe, but it's kind of weak. 00:30:02.720 --> 00:30:08.360 And then if you instead make it isoluminant, 00:30:08.360 --> 00:30:10.890 you get some in-between effect. 00:30:10.890 --> 00:30:16.170 So on the basis of that, we can conclude 00:30:16.170 --> 00:30:19.920 that very important in this display 00:30:19.920 --> 00:30:24.420 is that you have a gradual shift in the orientations 00:30:24.420 --> 00:30:26.220 of these squares. 00:30:26.220 --> 00:30:30.160 And that, in turn, suggests that somehow the way these squares 00:30:30.160 --> 00:30:36.230 activate orientation selective cells in V1 00:30:36.230 --> 00:30:38.670 plays an important role in giving rise 00:30:38.670 --> 00:30:40.775 to this strange illusory effect. 00:30:42.700 --> 00:30:43.200 OK. 00:30:43.200 --> 00:30:45.000 So now what we are going to do is 00:30:45.000 --> 00:30:48.090 to look at yet another illusory effect, which 00:30:48.090 --> 00:30:50.690 has a different source, and that one 00:30:50.690 --> 00:30:53.390 is the famous Muller-Lyer illusion. 00:30:53.390 --> 00:30:56.260 The fact is that these two bars here 00:30:56.260 --> 00:31:00.260 are identical in size-- do I have the picture-- like that. 00:31:00.260 --> 00:31:01.370 OK? 00:31:01.370 --> 00:31:05.420 But when you put these inducing elements, as we will call them, 00:31:05.420 --> 00:31:10.000 OK, on there, you'll swear that this one is longer 00:31:10.000 --> 00:31:11.110 than this one is. 00:31:11.110 --> 00:31:12.020 OK? 00:31:12.020 --> 00:31:16.040 So now we can ask the question, what kind of manipulations 00:31:16.040 --> 00:31:21.480 can we undertake to see what that is all about? 00:31:21.480 --> 00:31:24.060 And let me come back to that in a minute. 00:31:24.060 --> 00:31:26.090 So what you can do here is to make 00:31:26.090 --> 00:31:30.970 the inducing elements and the bars different in color 00:31:30.970 --> 00:31:35.510 or in contrast, and you still get a strong illusory effect. 00:31:35.510 --> 00:31:36.010 OK? 00:31:38.100 --> 00:31:42.350 So making them isoluminant or making them 00:31:42.350 --> 00:31:46.430 opposite in contrast does not eliminate this illusion. 00:31:46.430 --> 00:31:49.140 So the theory that somebody has come up with 00:31:49.140 --> 00:31:53.020 is shown here, which I'm not sure I buy, 00:31:53.020 --> 00:31:56.170 but that's nevertheless a theory, 00:31:56.170 --> 00:31:59.550 it claims that the brain interprets 00:31:59.550 --> 00:32:03.120 these two displays as differences in depth, 00:32:03.120 --> 00:32:08.910 as if here we had a protruding edge in a building inside 00:32:08.910 --> 00:32:12.210 or a receding edge as shown here. 00:32:12.210 --> 00:32:16.220 And that difference that we almost unconsciously, 00:32:16.220 --> 00:32:19.440 if you will, interpret as being in different depths 00:32:19.440 --> 00:32:22.200 is what gives rise to the illusion 00:32:22.200 --> 00:32:23.690 according to this theory. 00:32:24.980 --> 00:32:27.090 Of course, there's a simpler theory simply saying 00:32:27.090 --> 00:32:30.510 that we just integrate these two bits of information. 00:32:30.510 --> 00:32:31.010 All right. 00:32:31.010 --> 00:32:36.020 So now we can ask the question, if you have illusory effects 00:32:36.020 --> 00:32:40.070 of this sort, is this something that really takes place higher 00:32:40.070 --> 00:32:44.040 up in the brain, or does it take place in the retina where we 00:32:44.040 --> 00:32:49.910 cannot separate these inducing and testing elements? 00:32:49.910 --> 00:32:51.945 Well, the way you can do this-- I'm 00:32:51.945 --> 00:32:54.330 going to show you another illusion which 00:32:54.330 --> 00:32:55.450 is similar to this. 00:32:56.810 --> 00:32:59.520 Here we have the so-called Ebbinghaus illusion. 00:33:00.570 --> 00:33:02.800 This is a very similar kind of illusion. 00:33:02.800 --> 00:33:05.450 This is a test element and an inducing element. 00:33:05.450 --> 00:33:09.594 The test element here are these two circles, 00:33:09.594 --> 00:33:11.260 and you swear that this one is a smaller 00:33:11.260 --> 00:33:13.580 circle than that one is, right? 00:33:13.580 --> 00:33:15.590 And we have a bunch of small circles, 00:33:15.590 --> 00:33:17.630 and here we have bunch of large circles. 00:33:17.630 --> 00:33:19.920 So now we can ask, well, is this something 00:33:19.920 --> 00:33:22.280 that takes place in the retina or what? 00:33:22.280 --> 00:33:25.055 So what kind of experiment can you do? 00:33:25.055 --> 00:33:27.480 Well, the experiment that you can do 00:33:27.480 --> 00:33:30.045 is to present these displays interocularly. 00:33:31.870 --> 00:33:32.975 What does that mean? 00:33:32.975 --> 00:33:35.780 What that means is that you present this 00:33:35.780 --> 00:33:38.470 to the left eye and this to the right eye. 00:33:38.470 --> 00:33:41.430 And you present this device that's 00:33:41.430 --> 00:33:43.630 used by psychologists for this kind of thing. 00:33:43.630 --> 00:33:45.190 It's called a tachistoscope. 00:33:45.190 --> 00:33:46.120 You look through it. 00:33:46.120 --> 00:33:49.130 It can present things separately through the two eyes, 00:33:49.130 --> 00:33:51.190 and you can flash them on briefly 00:33:51.190 --> 00:33:54.120 so you don't get any binocular rivalry. 00:33:54.120 --> 00:33:56.660 And when you do that, then you can systematically 00:33:56.660 --> 00:34:01.080 vary the size of these to make sure 00:34:01.080 --> 00:34:04.150 that you can get a quantitative measure of it. 00:34:04.150 --> 00:34:06.860 And when you do that, you find that the illusory effect 00:34:06.860 --> 00:34:10.620 is just as good under interocular conditions 00:34:10.620 --> 00:34:13.430 as it is under binocular viewing conditions. 00:34:14.659 --> 00:34:16.469 Because of this then-- this has been 00:34:16.469 --> 00:34:18.790 done with many kinds of illusions 00:34:18.790 --> 00:34:20.580 including the Muller-Lyer one. 00:34:20.580 --> 00:34:24.580 And when you do that, you show, in essence, 00:34:24.580 --> 00:34:27.260 that this is an effect that takes place 00:34:27.260 --> 00:34:29.270 most likely in the cortex, certainly 00:34:29.270 --> 00:34:32.350 not in the retina or the lateral geniculate nucleus. 00:34:32.350 --> 00:34:34.639 It takes place someplace where the images 00:34:34.639 --> 00:34:36.210 from the left and two eyes converge. 00:34:37.290 --> 00:34:37.790 All right. 00:34:37.790 --> 00:34:43.139 So now yet another set of illusions 00:34:43.139 --> 00:34:45.679 that I want to show you. 00:34:47.230 --> 00:34:51.085 I think that these illusory effects are done. 00:34:52.280 --> 00:34:55.710 You can take this home and enjoy it or show it to your friends. 00:34:55.710 --> 00:34:59.200 Now I'm going to show you a set of just two 00:34:59.200 --> 00:35:01.140 kinds of motion illusions. 00:35:01.140 --> 00:35:02.590 OK, here's the first one. 00:35:03.910 --> 00:35:05.930 I want you to look at the screen, 00:35:05.930 --> 00:35:08.410 and I'm going to flash something on briefly, 00:35:08.410 --> 00:35:09.850 and I'm going to do it repeatedly. 00:35:09.850 --> 00:35:10.350 Ready? 00:35:14.420 --> 00:35:15.644 What do you see? 00:35:15.644 --> 00:35:16.810 Do you see something moving? 00:35:21.150 --> 00:35:24.370 A very brief sense of motion that you see. 00:35:24.370 --> 00:35:26.600 This twirls a little bit. 00:35:26.600 --> 00:35:31.480 And this is called the snake illusion, 00:35:31.480 --> 00:35:32.805 which is really curious. 00:35:34.540 --> 00:35:39.270 This has, again, appeared in quite a number of textbooks, 00:35:39.270 --> 00:35:41.900 and we do not have a really good explanation for it. 00:35:41.900 --> 00:35:45.720 But we have another one, which I think is related. 00:35:45.720 --> 00:35:46.220 OK. 00:35:46.220 --> 00:35:48.330 So what I'm going to do here, this 00:35:48.330 --> 00:35:51.150 is called the waterfall illusion. 00:35:51.150 --> 00:35:54.360 So what we are going to do here is 00:35:54.360 --> 00:35:56.230 I'm going to set this in motion. 00:35:56.230 --> 00:35:59.680 In this case, I don't want you to fixate. 00:35:59.680 --> 00:36:01.240 What I want you to do is just keep 00:36:01.240 --> 00:36:02.736 looking at it in various places. 00:36:02.736 --> 00:36:04.860 You can look here, here, here, and here, and so on. 00:36:04.860 --> 00:36:05.630 OK? 00:36:05.630 --> 00:36:07.740 And I'm going to let it run for a while, 00:36:07.740 --> 00:36:09.240 and then I'm going to stop it. 00:36:09.240 --> 00:36:09.740 OK? 00:36:09.740 --> 00:36:13.220 And I want you to tell me what is your aftereffect. 00:36:13.220 --> 00:36:13.720 Ready? 00:36:22.550 --> 00:36:23.050 OK. 00:36:23.050 --> 00:36:25.020 I'm about to stop it. 00:36:25.020 --> 00:36:25.754 Are you ready? 00:36:25.754 --> 00:36:27.170 Pay close attention to what you're 00:36:27.170 --> 00:36:28.545 going to perceive once it stops. 00:36:32.100 --> 00:36:33.630 Anybody see what happened? 00:36:33.630 --> 00:36:34.692 What did you see? 00:36:34.692 --> 00:36:36.080 AUDIENCE: [INAUDIBLE]. 00:36:36.080 --> 00:36:39.050 PROFESSOR: You see the water running up here, right? 00:36:39.050 --> 00:36:39.550 OK. 00:36:39.550 --> 00:36:42.550 I'll do it once more so you can appreciate it. 00:36:54.410 --> 00:36:55.831 Pretty nice, huh? 00:36:55.831 --> 00:36:56.330 OK. 00:36:56.330 --> 00:37:04.480 So this illusory effect is one that also occurs interocularly. 00:37:06.570 --> 00:37:14.210 And it has been shown that when you record from area MT, 00:37:14.210 --> 00:37:19.720 we talked about the fact that area MT is a region where cells 00:37:19.720 --> 00:37:24.360 respond selectively to direction of motion. 00:37:24.360 --> 00:37:26.690 Well, people have shown that if you, 00:37:26.690 --> 00:37:31.310 for an extended period of time, a minute maybe or whatever, 00:37:31.310 --> 00:37:39.590 you stimulate a cell in MT with this kind of a display, 00:37:39.590 --> 00:37:45.450 then after you stop it, OK, the cells with the opposite sense 00:37:45.450 --> 00:37:50.750 of motion will discharge some, which 00:37:50.750 --> 00:37:55.710 then is attributable to the upward sensation that you have. 00:37:55.710 --> 00:37:58.040 So it's clear enough that this is something 00:37:58.040 --> 00:38:01.690 that does take place predominantly by virtue 00:38:01.690 --> 00:38:06.140 of direction-selective cells and, probably, 00:38:06.140 --> 00:38:09.060 by direction-selective cells in area MT. 00:38:10.480 --> 00:38:10.980 All right. 00:38:10.980 --> 00:38:13.550 So then what we are going to do here 00:38:13.550 --> 00:38:15.750 is we're going to look at some other limitations 00:38:15.750 --> 00:38:19.080 and ambiguities in perception. 00:38:19.080 --> 00:38:21.190 And in particular, we're going to examine 00:38:21.190 --> 00:38:23.000 some figure/ground relationships. 00:38:24.310 --> 00:38:27.030 The first one I'm going to show you is this one. 00:38:27.030 --> 00:38:32.400 If you look at these two diagonals, OK, a and b, 00:38:32.400 --> 00:38:35.510 they certainly look very, very different, don't they? 00:38:35.510 --> 00:38:36.080 OK? 00:38:36.080 --> 00:38:38.260 One looks shorter and a different angle. 00:38:38.260 --> 00:38:41.857 But now if I eliminate the trapezoid in the surround, 00:38:41.857 --> 00:38:43.690 you can see that they're actually identical. 00:38:44.700 --> 00:38:45.200 OK? 00:38:45.200 --> 00:38:46.283 I'll show it to you again. 00:38:50.340 --> 00:38:51.660 Pretty dramatic. 00:38:51.660 --> 00:38:53.525 So that's an interesting illusory effect. 00:38:53.525 --> 00:38:55.380 And it's clear enough that this is something 00:38:55.380 --> 00:38:58.560 that takes place in higher cortical areas. 00:38:58.560 --> 00:39:02.430 Yet another one here that I'm going to show you also 00:39:02.430 --> 00:39:05.810 has something to do with figure/ground relationships, 00:39:05.810 --> 00:39:07.630 which has been extensively studied, 00:39:07.630 --> 00:39:09.640 especially by Gestalt psychologists. 00:39:09.640 --> 00:39:12.770 And I spoke about this a little bit the last time. 00:39:12.770 --> 00:39:16.890 What you see here is a bunch of vases, so to speak, right? 00:39:19.049 --> 00:39:19.840 That's good enough. 00:39:19.840 --> 00:39:21.170 So what I'm going to do now, I'm going 00:39:21.170 --> 00:39:22.544 to show you another set of vases. 00:39:22.544 --> 00:39:23.950 Are you ready? 00:39:23.950 --> 00:39:25.630 There is another set of vases. 00:39:25.630 --> 00:39:27.371 Everybody see the vases? 00:39:27.371 --> 00:39:28.870 Now, what if I tell you that there's 00:39:28.870 --> 00:39:30.380 much more to this than just vases? 00:39:31.440 --> 00:39:33.800 So if you look really closely, you're 00:39:33.800 --> 00:39:35.560 going to see something else there. 00:39:35.560 --> 00:39:37.017 What else do you see there? 00:39:37.017 --> 00:39:38.420 AUDIENCE: Faces. 00:39:38.420 --> 00:39:41.540 PROFESSOR: Oh, OK, two faces, two guys facing each other. 00:39:41.540 --> 00:39:47.520 Now, if I change the relative number of elements in here, 00:39:47.520 --> 00:39:50.130 which changes the figure/ground relationships, 00:39:50.130 --> 00:39:52.230 you get a different effect. 00:39:52.230 --> 00:39:54.650 Here you have-- predominantly you 00:39:54.650 --> 00:39:56.520 see the faces rather than the vases. 00:39:57.810 --> 00:39:58.620 OK? 00:39:58.620 --> 00:40:01.090 And that's a cute way of putting it. 00:40:01.090 --> 00:40:03.080 I see the faces, not the vases. 00:40:03.080 --> 00:40:04.580 And then you switch it, and you say, 00:40:04.580 --> 00:40:06.930 I see the vases, not the faces. 00:40:06.930 --> 00:40:07.530 All right? 00:40:07.530 --> 00:40:10.970 So that then is a way you can play around, again, 00:40:10.970 --> 00:40:14.050 with various illusory effects by fiddling around 00:40:14.050 --> 00:40:16.750 on your computer and creating things like that. 00:40:16.750 --> 00:40:17.510 OK. 00:40:17.510 --> 00:40:22.590 Now, yet another strange effect is 00:40:22.590 --> 00:40:25.330 where you can have difficulty telling what's 00:40:25.330 --> 00:40:27.250 what is this kind of a display. 00:40:27.250 --> 00:40:29.525 You say, well, what's happening here? 00:40:29.525 --> 00:40:31.310 Are arrows pointing to the right, 00:40:31.310 --> 00:40:33.370 or are they pointing to the left? 00:40:33.370 --> 00:40:37.570 And that reminds one of a fellow called Yogi Berra. 00:40:37.570 --> 00:40:40.140 How many of you have ever heard of Yogi Berra? 00:40:40.140 --> 00:40:41.840 Oh, my goodness, very good. 00:40:41.840 --> 00:40:42.340 OK. 00:40:42.340 --> 00:40:49.130 So he was a famous catcher for the New York Yankees, right? 00:40:49.130 --> 00:40:52.060 Well, he was a very interesting guy, 00:40:52.060 --> 00:40:54.170 had all kinds of interesting statements. 00:40:54.170 --> 00:40:58.010 And one of those many curious statements, one of them 00:40:58.010 --> 00:40:59.360 is-- it doesn't relate to this. 00:40:59.360 --> 00:41:00.680 It just comes into my head. 00:41:00.680 --> 00:41:02.810 "It isn't over until it's over." 00:41:02.810 --> 00:41:04.260 That's one of his, OK? 00:41:04.260 --> 00:41:06.080 This one here that's relevant is, 00:41:06.080 --> 00:41:09.163 "When you get to a fork in the road, take it." 00:41:09.163 --> 00:41:10.150 Yeah? 00:41:10.150 --> 00:41:11.810 And that's a Yogi Berra quote. 00:41:11.810 --> 00:41:14.290 And so what we have here is a sign that you 00:41:14.290 --> 00:41:17.629 come to a fork in the road, and you have to decide to take it. 00:41:17.629 --> 00:41:18.670 What are you going to do? 00:41:18.670 --> 00:41:21.253 Are you going to go this way, or are you going to go that way? 00:41:21.253 --> 00:41:23.090 So that's a kind of game you can play 00:41:23.090 --> 00:41:27.120 with this kind of strange effect that you can create 00:41:27.120 --> 00:41:31.490 that is confusing your ability to organize 00:41:31.490 --> 00:41:32.420 your visual percepts. 00:41:33.920 --> 00:41:37.350 Another famous one, maybe not that famous, actually. 00:41:37.350 --> 00:41:39.330 Maybe some of you have seen this. 00:41:39.330 --> 00:41:43.475 This, again, has appeared many, many years ago in The New. 00:41:43.475 --> 00:41:43.975 Yorker. 00:41:45.050 --> 00:41:47.030 It says, "I'm turning into my mother." 00:41:47.030 --> 00:41:48.580 Now, keep looking at it and tell me, 00:41:48.580 --> 00:41:51.980 why do you think that that comment is being made? 00:41:56.210 --> 00:42:00.730 What we have here, a young woman looking at another young woman, 00:42:00.730 --> 00:42:01.650 right? 00:42:01.650 --> 00:42:05.140 But now if you keep looking at this display, look. 00:42:05.140 --> 00:42:07.470 This is a nose here, and this is a mouth. 00:42:09.170 --> 00:42:10.450 You see an old lady now? 00:42:11.411 --> 00:42:11.910 OK. 00:42:13.190 --> 00:42:16.660 You played around with this, it gives you a confusing percept. 00:42:16.660 --> 00:42:18.820 And ultimately, you can see the young person 00:42:18.820 --> 00:42:22.840 looking that way or an old lady looking the other way. 00:42:22.840 --> 00:42:26.800 I mean looking not backwards but across. 00:42:27.942 --> 00:42:29.400 And that's why this statement says, 00:42:29.400 --> 00:42:33.350 "I'm turning into my mother." 00:42:33.350 --> 00:42:38.880 So the essence here is that playing around 00:42:38.880 --> 00:42:41.780 with these kinds of illusory effect is a lot of fun, 00:42:41.780 --> 00:42:44.940 and we all very much enjoy looking at illusions. 00:42:44.940 --> 00:42:47.340 And some of us like to play around with them 00:42:47.340 --> 00:42:52.875 to try to figure out what their existence is for the reasons. 00:42:52.875 --> 00:42:55.120 Now, I want to show you one more, 00:42:55.120 --> 00:42:57.720 which I think is very appropriate at this time. 00:42:57.720 --> 00:43:00.270 This is the famous Greek key motif. 00:43:00.270 --> 00:43:02.450 Everybody's seen this, I presume. 00:43:02.450 --> 00:43:05.895 Does everybody know where this is displayed, actually, 00:43:05.895 --> 00:43:07.265 at several places? 00:43:08.860 --> 00:43:10.041 In the Senate. 00:43:10.041 --> 00:43:10.540 OK? 00:43:10.540 --> 00:43:15.400 Whenever you have a picture of the Senate Building, 00:43:15.400 --> 00:43:16.660 you will see this. 00:43:16.660 --> 00:43:17.720 All right? 00:43:17.720 --> 00:43:20.190 Now, that's a little bit amusing, actually, 00:43:20.190 --> 00:43:21.350 if you think about it. 00:43:25.110 --> 00:43:27.460 So this has become symbolic of democracy. 00:43:28.930 --> 00:43:31.640 And considering just what happened most recently, 00:43:31.640 --> 00:43:34.210 you could almost say that, perhaps, 00:43:34.210 --> 00:43:39.240 what's happening here is that you cannot tell whether we see 00:43:39.240 --> 00:43:42.140 this guy or this guy here going around. 00:43:42.140 --> 00:43:42.640 OK? 00:43:43.457 --> 00:43:45.915 And one of them you could say, oh, these are the Democrats. 00:43:45.915 --> 00:43:47.730 These are the Republicans. 00:43:47.730 --> 00:43:51.066 And you have sort of a conflict as to which ones you see. 00:43:51.066 --> 00:43:51.800 All right? 00:43:51.800 --> 00:43:54.420 And that is, in a way, symbolic of what's going on today. 00:43:56.580 --> 00:43:59.120 But to educate you a little bit, which 00:43:59.120 --> 00:44:03.500 I suspect probably most of you know already anyway, 00:44:03.500 --> 00:44:09.930 this Greek key motif actually has its source 00:44:09.930 --> 00:44:13.080 in the so-called myth of the Labyrinth 00:44:13.080 --> 00:44:15.260 that had imprisoned the Minotaur. 00:44:16.810 --> 00:44:19.360 Who knows who the Minotaur is? 00:44:19.360 --> 00:44:20.240 Oh, very good. 00:44:20.240 --> 00:44:21.180 Many of you do. 00:44:21.180 --> 00:44:21.680 All right. 00:44:21.680 --> 00:44:28.190 So the Minotaur was a creature that 00:44:28.190 --> 00:44:31.410 was half bull and half human. 00:44:31.410 --> 00:44:32.140 All right? 00:44:32.140 --> 00:44:34.980 And what happened, according to this myth-- and let me tell you 00:44:34.980 --> 00:44:38.310 it's a myth, and it's a bit of risque thing. 00:44:38.310 --> 00:44:41.700 There was a king called King Minos 00:44:41.700 --> 00:44:46.580 whose wife Pasiphae-- Pasiphae, I'm not sure 00:44:46.580 --> 00:44:51.650 how you pronounce it properly-- is the one who had intercourse 00:44:51.650 --> 00:44:56.790 with a bull and gave birth-- sorry, 00:44:56.790 --> 00:44:59.700 had intercourse with a bull. 00:44:59.700 --> 00:45:02.270 And as a result of that, gave birth 00:45:02.270 --> 00:45:07.570 to the Minotaur, which was a half-bull, half-human monster, 00:45:07.570 --> 00:45:08.740 if you will. 00:45:08.740 --> 00:45:09.750 All right? 00:45:09.750 --> 00:45:13.510 And that became so threatening and so irritating, 00:45:13.510 --> 00:45:16.550 I guess, to King Minos that he asked 00:45:16.550 --> 00:45:24.610 one of his subjugates, Daedalus, to create a maze, a huge maze, 00:45:24.610 --> 00:45:28.300 and he put the Minotaur in it. 00:45:28.300 --> 00:45:30.900 So it was such a maze that the Minotaur just 00:45:30.900 --> 00:45:31.910 couldn't get out of it. 00:45:31.910 --> 00:45:32.880 All right? 00:45:32.880 --> 00:45:35.310 But then things got progressively worse, 00:45:35.310 --> 00:45:38.360 and King Minos decided that he had better 00:45:38.360 --> 00:45:42.220 get this Minotaur killed. 00:45:42.220 --> 00:45:50.979 And so he asked a person whose name is Theseus, OK, 00:45:50.979 --> 00:45:56.820 to go into the maze and kill the Minotaur. 00:45:58.420 --> 00:46:03.730 Now, the king also had another offspring 00:46:03.730 --> 00:46:06.450 whose name is Ariadne-- nice name, Ariadne. 00:46:08.740 --> 00:46:14.700 And she said to Theseus that if you take a ball of thread 00:46:14.700 --> 00:46:19.870 and you walk through the maze, once you 00:46:19.870 --> 00:46:24.070 do what you have to do, you can follow the thread back, 00:46:24.070 --> 00:46:26.470 and that way you can get out of the maze. 00:46:26.470 --> 00:46:29.120 So that's what Theseus did, and then he 00:46:29.120 --> 00:46:34.060 went on and killed the Minotaur. 00:46:35.890 --> 00:46:39.010 Now, the story didn't end there because somehow, 00:46:39.010 --> 00:46:43.890 for reasons I'm not sure I know the details anymore, 00:46:43.890 --> 00:46:47.420 King Minos had a fall out with Daedalus, 00:46:47.420 --> 00:46:51.760 and he jailed Daedalus and his son Icarus. 00:46:53.220 --> 00:46:54.620 And while they were in jail, they 00:46:54.620 --> 00:46:57.050 tried to figure out how they could escape. 00:46:57.050 --> 00:47:05.250 And so Daedalus fashioned a pair of wings 00:47:05.250 --> 00:47:11.170 made out of bird's feathers and wax, and then put it on Icarus. 00:47:11.170 --> 00:47:14.940 And then Icarus, by doing this, he could fly. 00:47:14.940 --> 00:47:15.780 OK? 00:47:15.780 --> 00:47:18.530 And then he flew and escaped from the prison. 00:47:18.530 --> 00:47:23.470 But unfortunately, he flew fairly high up 00:47:23.470 --> 00:47:25.180 where the sun was very hot. 00:47:25.180 --> 00:47:32.070 That melted the wax, and Icarus, sadly, fell to his death 00:47:32.070 --> 00:47:33.460 as a result. 00:47:33.460 --> 00:47:35.530 So that's quite the story. 00:47:35.530 --> 00:47:38.160 And that's something that obviously some of you 00:47:38.160 --> 00:47:38.940 already know. 00:47:38.940 --> 00:47:41.930 And those of you who don't, you will probably 00:47:41.930 --> 00:47:46.590 remember, as so many Greek myths are just fascinating, 00:47:46.590 --> 00:47:48.180 ancient, old stories. 00:47:48.180 --> 00:47:48.680 All right. 00:47:48.680 --> 00:47:53.000 So then here is actually a reconstruction 00:47:53.000 --> 00:47:59.760 of the kill of the Minotaur by Theseus. 00:47:59.760 --> 00:48:06.140 And this is actually a sculpture that exists in Greece. 00:48:08.410 --> 00:48:09.350 Pretty nice, huh? 00:48:11.630 --> 00:48:12.130 OK. 00:48:12.130 --> 00:48:17.960 Now, the last confusing image that I wanted to show you 00:48:17.960 --> 00:48:23.280 is this one that had appeared-- I think this appeared also 00:48:23.280 --> 00:48:28.160 in The New Yorker, but I'm not 100% sure. 00:48:28.160 --> 00:48:31.440 What you have here is a confusion-- 00:48:31.440 --> 00:48:34.660 do we have three or two prongs here? 00:48:34.660 --> 00:48:37.960 And the reason I show this sometimes in class 00:48:37.960 --> 00:48:42.960 is because in our department when it was originally 00:48:42.960 --> 00:48:45.930 formed by Hans-Lukas Teuber, he created 00:48:45.930 --> 00:48:47.930 what is called three prongs. 00:48:47.930 --> 00:48:48.990 OK? 00:48:48.990 --> 00:48:51.930 And then it became sort of an issue as to, well, 00:48:51.930 --> 00:48:54.020 what are the three prongs? 00:48:54.020 --> 00:48:55.020 Are there two prongs? 00:48:55.020 --> 00:48:56.120 Are there three prongs? 00:48:56.120 --> 00:48:57.810 Who belongs into which prong? 00:48:57.810 --> 00:49:00.430 And so somebody came up with this picture 00:49:00.430 --> 00:49:04.950 that was put up as a joke in the department when 00:49:04.950 --> 00:49:07.670 the department was still in one of the old buildings that 00:49:07.670 --> 00:49:09.771 doesn't even exist anymore. 00:49:09.771 --> 00:49:10.270 All right. 00:49:10.270 --> 00:49:17.500 So that then brings me to the end of the illusory effects 00:49:17.500 --> 00:49:20.280 that I thought we'd talk about briefly. 00:49:20.280 --> 00:49:23.645 And I will now move on and talk about visual prosthesis. 00:49:28.591 --> 00:49:29.090 All right. 00:49:29.090 --> 00:49:38.670 Now, as far as prosthetics are concerned, 00:49:38.670 --> 00:49:41.920 you're going to hear quite a bit about that from Chris Brown 00:49:41.920 --> 00:49:49.580 because in audition, an auditory prosthetic device, the cochlear 00:49:49.580 --> 00:49:52.400 implant, has become an incredible success. 00:49:53.420 --> 00:49:57.730 There are well over 50,000 people in the country now 00:49:57.730 --> 00:50:00.470 who have a prosthetic device like that 00:50:00.470 --> 00:50:03.940 that enables them to hear, to converse. 00:50:03.940 --> 00:50:05.340 It's really incredible. 00:50:05.340 --> 00:50:10.270 Unfortunately, when it comes to visual prostheses, 00:50:10.270 --> 00:50:14.180 we have nothing comparable at this stage. 00:50:14.180 --> 00:50:20.320 It's a much needed device that will take many, many, many more 00:50:20.320 --> 00:50:22.970 years to create. 00:50:22.970 --> 00:50:25.780 But I thought I'd give you sort of a crude sense of how 00:50:25.780 --> 00:50:28.690 that kind of work progresses because there are 00:50:28.690 --> 00:50:31.430 many laboratories that are trying to come up 00:50:31.430 --> 00:50:37.230 with something that will provide an ability for people to see. 00:50:37.230 --> 00:50:42.610 Now, I can tell you that in the world 00:50:42.610 --> 00:50:46.320 there are more than 40 million people who are blind. 00:50:46.320 --> 00:50:50.140 In the United States, there are more than a million individuals 00:50:50.140 --> 00:50:51.630 who are blind. 00:50:51.630 --> 00:50:54.706 So because of that-- that's a huge number-- 00:50:54.706 --> 00:50:57.920 it, of course, is extremely desirable to come up 00:50:57.920 --> 00:51:00.730 with some device to do this for you. 00:51:00.730 --> 00:51:04.040 Now, this is a very, very long history, 00:51:04.040 --> 00:51:05.815 and people have tried all sorts of things. 00:51:07.210 --> 00:51:13.250 One of those is to try to induce the ability to see something 00:51:13.250 --> 00:51:18.550 in the world that you normally see with your eye 00:51:18.550 --> 00:51:21.090 by putting it through some other sense. 00:51:21.090 --> 00:51:23.791 Now, a simple example of that is Braille. 00:51:23.791 --> 00:51:24.290 All right? 00:51:24.290 --> 00:51:26.270 Everybody knows what Braille is, right? 00:51:26.270 --> 00:51:32.260 That's you can use your hand to feel something that 00:51:32.260 --> 00:51:34.250 has protruding elements in it that 00:51:34.250 --> 00:51:36.290 can be shown in terms of letters. 00:51:36.290 --> 00:51:38.740 And so as you get into an elevator and you can't see, 00:51:38.740 --> 00:51:41.860 you can feel it, and then you can push the right button 00:51:41.860 --> 00:51:43.490 to get to the right floor. 00:51:43.490 --> 00:51:44.350 OK? 00:51:44.350 --> 00:51:46.430 Now, that's one way it has been done. 00:51:46.430 --> 00:51:51.480 Another approach had been to present stimulation 00:51:51.480 --> 00:51:55.020 to your somatosensory system, either by putting something 00:51:55.020 --> 00:51:58.220 on your back or putting something on your fingertip, 00:51:58.220 --> 00:52:03.220 that an image would be converted into actually activating 00:52:03.220 --> 00:52:06.070 your somatosensory system that would 00:52:06.070 --> 00:52:12.320 be equivalent to something that you would perceive if you could 00:52:12.320 --> 00:52:13.180 see. 00:52:13.180 --> 00:52:16.420 So that's the basic thing. 00:52:16.420 --> 00:52:19.650 And a tremendous amount of effort 00:52:19.650 --> 00:52:27.200 has been going on to try to come up with something to create 00:52:27.200 --> 00:52:30.140 a workable prosthetic device for the blind. 00:52:30.140 --> 00:52:31.590 Now, there's a lot of disagreement 00:52:31.590 --> 00:52:38.730 as to where such a device should be placed in the body 00:52:38.730 --> 00:52:42.940 and also a lot of debate of what kind of procedures 00:52:42.940 --> 00:52:43.750 should be used. 00:52:43.750 --> 00:52:46.920 Now, some people advocate that it 00:52:46.920 --> 00:52:49.200 should be done on the retina. 00:52:49.200 --> 00:52:51.140 Some people advocate it should be 00:52:51.140 --> 00:52:55.890 done in the lateral geniculate nucleus. 00:52:55.890 --> 00:52:57.780 Actually, some work is being done 00:52:57.780 --> 00:53:01.200 at Harvard trying to come up with something 00:53:01.200 --> 00:53:02.860 for the lateral geniculate nucleus. 00:53:02.860 --> 00:53:05.355 And there's some people who think 00:53:05.355 --> 00:53:08.940 it would be best to do it in the visual cortex. 00:53:08.940 --> 00:53:12.290 So let me sort of go into this in a bit more detail. 00:53:12.290 --> 00:53:15.920 And let me, first of all, raise the question of, 00:53:15.920 --> 00:53:19.630 what kinds of devices, and what kinds of necessities 00:53:19.630 --> 00:53:23.920 do we have for converting vision into something 00:53:23.920 --> 00:53:27.080 that enables blind people to see? 00:53:27.080 --> 00:53:29.840 Well, first of all, it's very important 00:53:29.840 --> 00:53:33.210 to be able to see some basic patterns, very important to see 00:53:33.210 --> 00:53:36.380 motion, and very important, often ignored, 00:53:36.380 --> 00:53:38.230 is that you got to be able to see something 00:53:38.230 --> 00:53:41.300 in the third dimension, because a person who is blind, 00:53:41.300 --> 00:53:43.630 he should be able to walk in the world. 00:53:43.630 --> 00:53:45.330 And if you're going to walk around, 00:53:45.330 --> 00:53:48.740 you should be able to see where things are in depth. 00:53:48.740 --> 00:53:51.060 So those are the basic requirements. 00:53:51.060 --> 00:53:55.600 You're certainly not required to process color information. 00:53:55.600 --> 00:53:56.601 So that's not important. 00:53:56.601 --> 00:53:59.016 And there are several other things that are not important. 00:53:59.016 --> 00:54:00.880 But these are the three most important ones 00:54:00.880 --> 00:54:03.570 that you would like to have if you are going 00:54:03.570 --> 00:54:05.940 to create a prosthetics device for the blind. 00:54:05.940 --> 00:54:08.215 Now, then the next question is, how are we 00:54:08.215 --> 00:54:09.840 going to deal with this, and what kinds 00:54:09.840 --> 00:54:12.080 of issues and problems are we going to deal with? 00:54:12.080 --> 00:54:16.520 First of all, the big issue is what kind of prosthetic device 00:54:16.520 --> 00:54:17.255 should we use. 00:54:18.810 --> 00:54:21.040 One of those that I've already mentioned 00:54:21.040 --> 00:54:26.670 is to try to convert visual impressions 00:54:26.670 --> 00:54:31.580 into other modalities such as your somatosensory system. 00:54:31.580 --> 00:54:32.320 All right? 00:54:32.320 --> 00:54:39.260 The other is to create some sort of stimulating device, 00:54:39.260 --> 00:54:42.280 which there are various kinds. 00:54:42.280 --> 00:54:44.130 But one of those would be, for example, 00:54:44.130 --> 00:54:47.500 to put a bunch of electrodes into the brain, 00:54:47.500 --> 00:54:52.210 or into the retina, or whatever, and then selectively stimulate 00:54:52.210 --> 00:54:53.950 through this series of electrodes 00:54:53.950 --> 00:54:56.780 to mimic the visual field. 00:54:56.780 --> 00:54:59.550 Now, the other one, and it is a big issue, 00:54:59.550 --> 00:55:03.500 what brain areas should be considered? 00:55:03.500 --> 00:55:06.110 And because we're at an early stage in this, 00:55:06.110 --> 00:55:11.000 I think it is desirable for people 00:55:11.000 --> 00:55:17.390 to pursue just about any area they think might work. 00:55:18.400 --> 00:55:23.859 Now, the problem with a retinal implant-- 00:55:23.859 --> 00:55:25.150 there are two problems with it. 00:55:25.150 --> 00:55:29.360 One is the retina is extremely small, and very confined, 00:55:29.360 --> 00:55:36.630 and has these millions of receptors and ganglion cells, 00:55:36.630 --> 00:55:39.180 and how can you selectively activate them? 00:55:39.180 --> 00:55:41.000 Now, the biggest problem, though, 00:55:41.000 --> 00:55:46.060 is that in most blind people, the retina 00:55:46.060 --> 00:55:48.490 becomes non-functional. 00:55:48.490 --> 00:55:53.380 And even worse, it degenerates in a relatively short time 00:55:53.380 --> 00:55:58.225 so that it no longer is useful for electrical stimulation 00:55:58.225 --> 00:56:00.540 or for whatever kind of stimulation 00:56:00.540 --> 00:56:02.340 you wish to engage in. 00:56:02.340 --> 00:56:05.549 So because of that, some people have 00:56:05.549 --> 00:56:07.340 moved on to the lateral geniculate nucleus. 00:56:08.590 --> 00:56:12.440 That area has some problems because it's deep in the brain, 00:56:12.440 --> 00:56:13.465 and it's quite small. 00:56:14.930 --> 00:56:15.920 OK? 00:56:15.920 --> 00:56:19.370 And also, it tends to degenerate over time, 00:56:19.370 --> 00:56:24.200 although it takes it a lot longer than the degeneration 00:56:24.200 --> 00:56:25.460 of the retina. 00:56:25.460 --> 00:56:31.305 Thirdly, people have considered stimulating the visual cortex, 00:56:31.305 --> 00:56:32.680 and that's what I'm going to talk 00:56:32.680 --> 00:56:35.110 about in a bit more detail just now. 00:56:36.470 --> 00:56:39.680 But the last thing that's very important here-- I think it's 00:56:39.680 --> 00:56:40.340 the last thing. 00:56:40.340 --> 00:56:41.798 Yeah, just let me go back. 00:56:46.658 --> 00:56:47.610 Ah, sorry. 00:56:52.880 --> 00:56:55.510 The third big issue is, can you create 00:56:55.510 --> 00:56:57.275 a device that has longevity? 00:56:58.440 --> 00:57:00.350 As soon as you put something into the brain 00:57:00.350 --> 00:57:03.350 and into the eye, there's going to be 00:57:03.350 --> 00:57:05.520 some sort of adverse reaction that 00:57:05.520 --> 00:57:07.830 may kill the cells at the tip of the electrodes 00:57:07.830 --> 00:57:08.950 or something like that. 00:57:08.950 --> 00:57:12.450 So all kinds of approaches have been 00:57:12.450 --> 00:57:19.400 created to try to increase the longevity of the device. 00:57:19.400 --> 00:57:21.770 Now, I'm going to next look at what 00:57:21.770 --> 00:57:24.990 happens in the visual cortex, which 00:57:24.990 --> 00:57:29.750 is a promising, at least I think it is, a promising site 00:57:29.750 --> 00:57:33.810 where you could electrically stimulate and use 00:57:33.810 --> 00:57:37.310 that to create a visual impression. 00:57:37.310 --> 00:57:37.810 All right. 00:57:37.810 --> 00:57:40.240 If you do that-- here is a monkey brain. 00:57:40.240 --> 00:57:42.030 The monkey is very well suited for this 00:57:42.030 --> 00:57:44.180 because, as I have mentioned before, the posterior 00:57:44.180 --> 00:57:47.330 part of the cortex-- see, this is area V1-- is lissencephalic. 00:57:48.500 --> 00:57:54.720 And so it's relatively easy to place accurately electrodes 00:57:54.720 --> 00:57:55.690 into it. 00:57:55.690 --> 00:57:56.340 OK? 00:57:56.340 --> 00:58:00.110 So then if what you do here-- here, just a reminder, 00:58:00.110 --> 00:58:01.510 is the central sulcus. 00:58:01.510 --> 00:58:02.810 Here's the lunate sulcus. 00:58:02.810 --> 00:58:04.860 And here, of course, is area V1. 00:58:04.860 --> 00:58:07.730 If you now take the contralateral hemifield, 00:58:07.730 --> 00:58:10.930 that projects into this region. 00:58:10.930 --> 00:58:11.430 OK? 00:58:11.430 --> 00:58:13.670 This is shown out to about five degrees. 00:58:13.670 --> 00:58:16.220 That will go to around here like that. 00:58:16.220 --> 00:58:17.510 OK? 00:58:17.510 --> 00:58:21.050 So now the next big question that, initially, 00:58:21.050 --> 00:58:23.080 had been done actually in humans, 00:58:23.080 --> 00:58:30.570 what do you see when you insert a microelectrode 00:58:30.570 --> 00:58:34.490 into the brain-- as I say, this was done on humans initially-- 00:58:34.490 --> 00:58:37.210 and you electrically stimulate? 00:58:37.210 --> 00:58:38.030 What do you see? 00:58:38.030 --> 00:58:42.310 And this work was done by Brindley many years ago. 00:58:42.310 --> 00:58:46.730 And he discovered that when you electrically 00:58:46.730 --> 00:58:54.220 stimulate in the human primary visual cortex, what you get 00:58:54.220 --> 00:58:58.090 is a small, star-like image. 00:58:58.090 --> 00:58:59.280 All right? 00:58:59.280 --> 00:59:00.700 So that's what he found. 00:59:00.700 --> 00:59:06.150 And so that then triggered a great deal of research, 00:59:06.150 --> 00:59:10.080 including work in monkeys, to determine 00:59:10.080 --> 00:59:13.160 what could be done to create a prosthetic device that 00:59:13.160 --> 00:59:14.610 might just work. 00:59:14.610 --> 00:59:17.870 So the first step then was to try 00:59:17.870 --> 00:59:19.880 to determine what happens when you 00:59:19.880 --> 00:59:21.530 put a microelectrode in here. 00:59:21.530 --> 00:59:26.560 This is, by the way, the layout of the visual field 00:59:26.560 --> 00:59:29.750 in the posterior cortex with much more area devoted 00:59:29.750 --> 00:59:33.550 to the fovea than into the periphery because, of course, 00:59:33.550 --> 00:59:39.160 there's a much higher packing density of receptor cells 00:59:39.160 --> 00:59:41.850 as well as retinal ganglion cells in the fovea 00:59:41.850 --> 00:59:44.470 representation than further out. 00:59:44.470 --> 00:59:47.860 In the cortex, the thickness of the gray matter is constant. 00:59:47.860 --> 00:59:50.490 It's about roughly 2 millimeters thick. 00:59:52.330 --> 00:59:57.120 And the frequency, spatial frequency of the cells, 00:59:57.120 --> 00:59:59.100 is pretty constant as well. 00:59:59.100 --> 01:00:03.330 So therefore, you need to allocate more space 01:00:03.330 --> 01:00:09.220 to the incoming huge number of retinal ganglion 01:00:09.220 --> 01:00:11.640 cells projected to the geniculate 01:00:11.640 --> 01:00:14.405 and then to the cortex than for the periphery. 01:00:14.405 --> 01:00:15.780 And that's why you have this kind 01:00:15.780 --> 01:00:18.605 of a layout of much more space allocated 01:00:18.605 --> 01:00:21.440 to the central than to peripheral vision. 01:00:21.440 --> 01:00:24.030 So that's a very important factor 01:00:24.030 --> 01:00:27.870 in considering how to create a prosthetic device. 01:00:27.870 --> 01:00:32.495 So now the first step in doing this kind of stuff, 01:00:32.495 --> 01:00:36.090 if you're going to do serious experiments on monkeys, 01:00:36.090 --> 01:00:38.630 is to determine, what does the monkey 01:00:38.630 --> 01:00:42.590 see when you electrically stimulate the visual cortex? 01:00:42.590 --> 01:00:46.130 And so the experiment that's been done here 01:00:46.130 --> 01:00:50.050 is one in which a monkey has been trained 01:00:50.050 --> 01:00:58.040 to make an eye movement to the bigger of two visual stimuli 01:00:58.040 --> 01:00:59.750 or the brighter of two stimuli. 01:00:59.750 --> 01:01:02.540 In this case, we have two visual targets. 01:01:02.540 --> 01:01:04.050 By the way, one of these is going 01:01:04.050 --> 01:01:08.600 to appear where the receptive field is of the cells 01:01:08.600 --> 01:01:10.884 that you're going to stimulate in the visual cortex. 01:01:10.884 --> 01:01:13.300 So then you may sometimes make this one brighter, this one 01:01:13.300 --> 01:01:13.990 brighter. 01:01:13.990 --> 01:01:16.870 And the monkey makes a saccade to it, 01:01:16.870 --> 01:01:19.000 and he gets a drop of apple juice for reward. 01:01:19.000 --> 01:01:22.030 And then you can systematically vary the relative size 01:01:22.030 --> 01:01:25.750 of the two stimuli to see what kind of a function you get. 01:01:25.750 --> 01:01:29.090 And the same thing you can do for the relative brightness 01:01:29.090 --> 01:01:29.740 of it. 01:01:29.740 --> 01:01:33.510 And then on some trials interspersed with those trials, 01:01:33.510 --> 01:01:35.870 you present [INAUDIBLE] stimulation 01:01:35.870 --> 01:01:37.850 with the visual target and see which 01:01:37.850 --> 01:01:39.410 one does the monkey go to. 01:01:39.410 --> 01:01:41.210 And then you systematically vary, 01:01:41.210 --> 01:01:45.140 as I said, the brightness or the size of those stimuli. 01:01:45.140 --> 01:01:45.640 OK? 01:01:47.140 --> 01:01:50.270 We vary the contrast in this case, and in this case, 01:01:50.270 --> 01:01:51.510 we vary the size. 01:01:51.510 --> 01:01:52.529 OK? 01:01:52.529 --> 01:01:53.820 So that's the basic experiment. 01:01:53.820 --> 01:01:55.880 And if you do that systematically, 01:01:55.880 --> 01:02:02.600 you can tell exactly what the size and the contrast 01:02:02.600 --> 01:02:07.070 is of the visual stimulus created 01:02:07.070 --> 01:02:08.470 by electrical stimulation. 01:02:08.470 --> 01:02:10.900 If you do that-- here's an example. 01:02:10.900 --> 01:02:13.140 When you present two visual stimuli 01:02:13.140 --> 01:02:18.500 and you vary the percent contrast difference, all right, 01:02:18.500 --> 01:02:21.630 of the two stimuli, you can see that it crosses over, indeed, 01:02:21.630 --> 01:02:22.850 when they're identical. 01:02:22.850 --> 01:02:25.270 So that tells you that this system works. 01:02:25.270 --> 01:02:27.120 It says these are identical. 01:02:27.120 --> 01:02:29.155 And the same thing is true for the size series. 01:02:30.320 --> 01:02:32.820 So this way then, if you do the same experiment 01:02:32.820 --> 01:02:36.620 with electrical stimulation, you can find out 01:02:36.620 --> 01:02:38.940 where this 50% crossover point is. 01:02:38.940 --> 01:02:42.050 This is for contrast, and this is for size. 01:02:42.050 --> 01:02:45.340 So then you can establish this with slightly different 01:02:45.340 --> 01:02:47.270 locations in the visual field. 01:02:47.270 --> 01:02:49.440 And if you do that overall, you come up 01:02:49.440 --> 01:02:51.850 with a summary statement, which says 01:02:51.850 --> 01:02:55.380 that when you use currents between 20 and 120 microamps, 01:02:55.380 --> 01:02:59.130 OK, at eccentricities between 2 and 1/2 and 3 and 1/2 degrees, 01:02:59.130 --> 01:03:02.190 the contrast of the visual percept created in monkeys 01:03:02.190 --> 01:03:06.280 is 6% to 12%, and the size is between 15 and 20 minutes 01:03:06.280 --> 01:03:07.880 of visual angle. 01:03:07.880 --> 01:03:10.710 Now, that's quite small fortunately. 01:03:10.710 --> 01:03:15.010 So it's a local, little spot, like a star-like image. 01:03:15.010 --> 01:03:18.980 Now, what do we mean by 6% to 12% contrast? 01:03:18.980 --> 01:03:20.712 So let me tell you this, that that's 01:03:20.712 --> 01:03:22.170 a useful thing for you to memorize. 01:03:23.200 --> 01:03:25.400 We talk about percent contrast. 01:03:25.400 --> 01:03:28.560 What you do is you measure, for example, 01:03:28.560 --> 01:03:31.490 in candelas per square meter, what the contrast is 01:03:31.490 --> 01:03:33.800 of this and what is the contest of the background. 01:03:33.800 --> 01:03:36.710 Then you subtract one from the other, divide it 01:03:36.710 --> 01:03:39.580 by the sum of the two, and you multiply by 100. 01:03:39.580 --> 01:03:42.490 So when you do that, this is roughly 8% contrast. 01:03:42.490 --> 01:03:45.070 This is roughly 75% contrast. 01:03:45.070 --> 01:03:48.060 So it gives you a sense of what contrast means. 01:03:48.060 --> 01:03:48.640 All right? 01:03:48.640 --> 01:03:52.350 So what that means, in essence, that the visual stimulus 01:03:52.350 --> 01:03:54.065 created by the electrical stimulation 01:03:54.065 --> 01:03:56.030 is fairly low contrast. 01:03:56.030 --> 01:03:56.530 All right. 01:03:56.530 --> 01:04:00.780 So now the next important factor, 01:04:00.780 --> 01:04:04.120 once one knows what the visual percept is 01:04:04.120 --> 01:04:06.840 that is created by electrical stimulation in the monkey, 01:04:06.840 --> 01:04:08.890 the next thing is to ask the question, well, 01:04:08.890 --> 01:04:12.270 what kind of electrode array should one put together? 01:04:18.210 --> 01:04:18.840 All right. 01:04:18.840 --> 01:04:21.540 Well, the big issue here has to do 01:04:21.540 --> 01:04:23.900 with the so-called magnification factor 01:04:23.900 --> 01:04:25.220 that I just talked about. 01:04:25.220 --> 01:04:26.816 So let's look at this analytically. 01:04:26.816 --> 01:04:28.440 First of all, I already mentioned this. 01:04:28.440 --> 01:04:32.750 We have a very high packing density of photoreceptors 01:04:32.750 --> 01:04:35.720 in the fovea and much lower density further out. 01:04:36.930 --> 01:04:41.660 And so what you have is that this plate 01:04:41.660 --> 01:04:45.365 goes through the optic nerve up to the visual cortex 01:04:45.365 --> 01:04:47.580 to the lateral geniculate nucleus. 01:04:47.580 --> 01:04:49.710 So that's the basic arrangement, meaning 01:04:49.710 --> 01:04:53.590 that there are many, many more retinal ganglion 01:04:53.590 --> 01:04:58.220 cells per unit area that impinge on the visual cortex 01:04:58.220 --> 01:05:00.490 near the fovea than in the periphery. 01:05:01.550 --> 01:05:06.150 So then to examine this, here we have, again, another picture. 01:05:06.150 --> 01:05:07.560 It's the geniculate again. 01:05:07.560 --> 01:05:11.370 Here's the fact that the cortex is constant in thickness. 01:05:11.370 --> 01:05:15.810 And here's the layout of the visual field on the posterior 01:05:15.810 --> 01:05:20.210 cortex with more area allocated to the fovea 01:05:20.210 --> 01:05:22.300 than to-- excuse me, than to periphery. 01:05:22.300 --> 01:05:22.800 All right. 01:05:22.800 --> 01:05:25.455 So now I've shown you a variant of this before. 01:05:26.640 --> 01:05:31.730 If you present an arrow across the visual field here, 01:05:31.730 --> 01:05:36.470 this is the impression made with much more activation 01:05:36.470 --> 01:05:39.630 in central vision than in peripheral vision. 01:05:39.630 --> 01:05:41.620 I mean, you don't have to worry about that. 01:05:41.620 --> 01:05:45.650 That simply tells you, as we had discussed before, 01:05:45.650 --> 01:05:49.250 that the images created depend heavily 01:05:49.250 --> 01:05:52.700 on where you stimulate in the visual field. 01:05:52.700 --> 01:05:53.210 OK. 01:05:53.210 --> 01:05:56.800 So now to look at this quantitatively, 01:05:56.800 --> 01:06:00.200 so we put a bunch of dots up in the visual field. 01:06:00.200 --> 01:06:02.370 This is in the central four degrees, 01:06:02.370 --> 01:06:05.340 and this is the actual activation 01:06:05.340 --> 01:06:06.680 in the visual cortex. 01:06:06.680 --> 01:06:07.180 OK? 01:06:08.400 --> 01:06:12.940 Meaning that in the fovea, we have two elements here, 01:06:12.940 --> 01:06:15.990 this one and this one, here and here, that they take up 01:06:15.990 --> 01:06:18.720 a lot of space, whereas in the periphery, 01:06:18.720 --> 01:06:20.480 OK, they're very close to each other. 01:06:20.480 --> 01:06:23.810 That's due to the so-called magnification factor. 01:06:23.810 --> 01:06:26.310 And so one of the basic rules is that we've 01:06:26.310 --> 01:06:32.190 got to be able to stimulate the visual cortex taking 01:06:32.190 --> 01:06:34.620 into account the magnification factor. 01:06:34.620 --> 01:06:37.310 So if you have this kind of spatial arrangement, 01:06:37.310 --> 01:06:39.360 if we then were to put a bunch of electrodes 01:06:39.360 --> 01:06:43.290 in like this, which we will call proportional electrodes, 01:06:43.290 --> 01:06:45.640 then if you stimulated all those, 01:06:45.640 --> 01:06:49.950 you would get a similar square with one proviso, 01:06:49.950 --> 01:06:53.276 namely that the little dots would be smaller in the fovea, 01:06:53.276 --> 01:06:54.650 obviously, than in the periphery. 01:06:55.740 --> 01:06:59.350 So then if you have this kind of arrangement, what you can do 01:06:59.350 --> 01:07:06.410 is you can take a camera that has a fixed number of elements, 01:07:06.410 --> 01:07:08.480 and each of those elements you hook up 01:07:08.480 --> 01:07:09.960 to one of the electrodes. 01:07:09.960 --> 01:07:12.390 If you put the word "fiat lux" there, 01:07:12.390 --> 01:07:16.120 this is what you hypothetically would create, like that. 01:07:16.120 --> 01:07:16.960 OK? 01:07:16.960 --> 01:07:18.550 That's what it would look like. 01:07:18.550 --> 01:07:22.620 So in contrast with this arrangement, 01:07:22.620 --> 01:07:25.300 if we were to use an electrode array, 01:07:25.300 --> 01:07:28.070 which you can readily buy-- you cannot buy proportional arrays. 01:07:28.070 --> 01:07:30.100 You have to somehow create them yourself. 01:07:30.100 --> 01:07:35.280 But if you buy a fixed array, which the elements are 01:07:35.280 --> 01:07:37.740 equidistant, then if you stimulated all those 01:07:37.740 --> 01:07:41.840 you would get a butterfly image, butterfly image. 01:07:41.840 --> 01:07:43.920 I will show a movie of that as well. 01:07:43.920 --> 01:07:47.370 And that means that if you do that same conversion as before, 01:07:47.370 --> 01:07:50.860 this is what you would see in contrast to this. 01:07:50.860 --> 01:07:54.170 So in other words, it's a greatly distorted vision 01:07:54.170 --> 01:07:57.120 because you didn't take into account the magnification 01:07:57.120 --> 01:07:58.061 factor. 01:07:58.061 --> 01:07:58.560 All right. 01:07:58.560 --> 01:08:01.390 So then what you can do, actually, you 01:08:01.390 --> 01:08:05.603 can also increase the number of electrodes in the center here. 01:08:05.603 --> 01:08:06.610 OK? 01:08:06.610 --> 01:08:08.700 And that would mean-- in this case 01:08:08.700 --> 01:08:13.360 we have 428 elements-- that you can recreate fiat lux pretty 01:08:13.360 --> 01:08:17.029 well with this increased number of elements, 01:08:17.029 --> 01:08:20.750 where with the original set you couldn't make anything out. 01:08:20.750 --> 01:08:23.020 Now, the alternate to this would be 01:08:23.020 --> 01:08:24.420 that you could take a whole bunch 01:08:24.420 --> 01:08:28.069 of equally-spaced electrodes, which you can commercially buy, 01:08:28.069 --> 01:08:29.140 and put them in. 01:08:29.140 --> 01:08:33.140 But then to get the same acuity, you would need, in this case, 01:08:33.140 --> 01:08:37.130 something like over 600 electrodes, 01:08:37.130 --> 01:08:40.950 whereas in this case you would only need about 420. 01:08:40.950 --> 01:08:43.270 So because of that, obviously we want 01:08:43.270 --> 01:08:46.240 to minimize the number of things you put into the brain, 01:08:46.240 --> 01:08:49.670 it's desirable to use a proportional device. 01:08:49.670 --> 01:08:51.279 So that's the first big step. 01:08:51.279 --> 01:08:54.689 Now, let's assume that you have created a proportional device. 01:08:54.689 --> 01:08:58.810 The next thing would be to see, how do you stimulate these 01:08:58.810 --> 01:09:02.930 through a camera? 01:09:02.930 --> 01:09:08.460 So what you do then is an experiment 01:09:08.460 --> 01:09:10.729 in which you take a human subject, 01:09:10.729 --> 01:09:13.590 and you can put a camera on a helmet, 01:09:13.590 --> 01:09:16.649 and also, the person looks at a monitor. 01:09:16.649 --> 01:09:18.865 And otherwise, the rest of the world 01:09:18.865 --> 01:09:22.020 is excluded by having this bezel here. 01:09:22.020 --> 01:09:24.670 And then the camera looks at something, 01:09:24.670 --> 01:09:28.020 and then it can display the image here. 01:09:28.020 --> 01:09:31.460 Well, one thing it can do, it can display the image as it is. 01:09:31.460 --> 01:09:33.590 But the other thing that's much more important, 01:09:33.590 --> 01:09:38.740 it can display that image as if electrical stimulation 01:09:38.740 --> 01:09:39.950 had taken place. 01:09:39.950 --> 01:09:42.109 Now, this becomes complicated because you 01:09:42.109 --> 01:09:53.490 have to do a lot of conversion of the input from the camera 01:09:53.490 --> 01:09:58.780 to the computer that's going to then drive the electrodes. 01:09:58.780 --> 01:10:01.990 Now, the first important point, I mentioned this before, 01:10:01.990 --> 01:10:06.690 is that if you present a small spot-- this is actual real data 01:10:06.690 --> 01:10:12.020 here-- you present a small spot in the center of a small array 01:10:12.020 --> 01:10:16.900 of cells in the visual cortex, you get a vigorous response, 01:10:16.900 --> 01:10:18.950 whether a light increment or light decrement. 01:10:18.950 --> 01:10:22.120 But if you use a larger spot, you get no response, 01:10:22.120 --> 01:10:26.140 meaning that the cortical cells look at local differences. 01:10:26.140 --> 01:10:28.620 That has a lot to do with what we talked about 01:10:28.620 --> 01:10:29.950 with the illusions. 01:10:29.950 --> 01:10:32.140 And so what you have to create is 01:10:32.140 --> 01:10:35.050 an algorithm that mimics that. 01:10:35.050 --> 01:10:36.910 And so the way you do that is you 01:10:36.910 --> 01:10:40.770 have these individual elements for each of the electrodes. 01:10:40.770 --> 01:10:48.250 And if that individual element in the computer that 01:10:48.250 --> 01:10:52.650 gets the input from the camera, if the whole element gets 01:10:52.650 --> 01:10:57.480 activated, there's no real activation of the electrode. 01:10:57.480 --> 01:10:59.210 But if an edge appears like here, 01:10:59.210 --> 01:11:02.570 here, or here, then the electrode is activated. 01:11:02.570 --> 01:11:03.190 OK? 01:11:03.190 --> 01:11:07.830 So you create local differences by this conversion system, 01:11:07.830 --> 01:11:12.160 all right, by this algorithm that then, 01:11:12.160 --> 01:11:13.900 when the camera looks at the world, 01:11:13.900 --> 01:11:17.000 will respond whenever there's an edge, an illumination 01:11:17.000 --> 01:11:17.751 difference. 01:11:17.751 --> 01:11:18.250 OK. 01:11:18.250 --> 01:11:22.160 So now the way this looks is like this. 01:11:23.440 --> 01:11:26.295 Again, as I've told you, whatever is in the fovea, 01:11:26.295 --> 01:11:29.320 the dots are smaller and further out, but that's OK. 01:11:29.320 --> 01:11:31.820 So that's the basic activation system. 01:11:31.820 --> 01:11:34.530 And so now what we are going to do, 01:11:34.530 --> 01:11:39.960 having created such an algorithm, 01:11:39.960 --> 01:11:44.660 we're going to examine what the person who 01:11:44.660 --> 01:11:48.030 carries this camera on his head can actually see. 01:11:48.030 --> 01:11:50.630 Now, the way this can be done, you can either physically 01:11:50.630 --> 01:11:54.223 move the stimuli in the world, or the person looks at it 01:11:54.223 --> 01:11:56.820 and can slowly move his head around. 01:11:56.820 --> 01:11:57.750 OK? 01:11:57.750 --> 01:12:10.170 So let's first look at what would 01:12:10.170 --> 01:12:19.790 happen if, instead of a proportional display, 01:12:19.790 --> 01:12:24.490 we used an equally-spaced display like that butterfly 01:12:24.490 --> 01:12:25.040 display. 01:12:25.040 --> 01:12:25.640 OK? 01:12:25.640 --> 01:12:28.720 So what I'm going to show you here on the left 01:12:28.720 --> 01:12:34.795 is what the camera is viewing. 01:12:35.820 --> 01:12:38.560 And here I'm going to show you what is actually 01:12:38.560 --> 01:12:43.510 perceived by this algorithm, by the person, where this image is 01:12:43.510 --> 01:12:46.320 converted into what he's looking at through that bezel. 01:12:46.320 --> 01:12:47.076 OK, you ready? 01:12:57.300 --> 01:12:59.330 I mean, you don't see a square at all, right? 01:12:59.330 --> 01:13:02.930 What you see is, because you're not taking into account 01:13:02.930 --> 01:13:05.430 the magnification factor, you get a huge distortion. 01:13:06.480 --> 01:13:10.290 So now let's go move on and ask the question, 01:13:10.290 --> 01:13:17.540 well, what happens when you are using a proportional display? 01:13:17.540 --> 01:13:18.040 All right? 01:13:23.440 --> 01:13:25.845 And now we get a rather nice story. 01:13:29.180 --> 01:13:30.920 Again, the visual display's on the left, 01:13:30.920 --> 01:13:32.605 and the image created is on the right. 01:13:32.605 --> 01:13:33.104 Ready? 01:13:37.280 --> 01:13:41.750 So the rule, therefore, is that you 01:13:41.750 --> 01:13:51.480 must use a proportional image, proportional display, that 01:13:51.480 --> 01:13:54.790 can be done either by having it truly proportional as you put 01:13:54.790 --> 01:14:00.030 it into the visual cortex or by having an equally-spaced array, 01:14:00.030 --> 01:14:04.250 and the computer computes the proportional arrangement. 01:14:04.250 --> 01:14:05.670 But the disadvantage of the latter 01:14:05.670 --> 01:14:07.503 is that you need many, many more electrodes. 01:14:08.120 --> 01:14:08.620 All right. 01:14:08.620 --> 01:14:11.360 Now, what I'm going to do is I'm going 01:14:11.360 --> 01:14:17.310 to show you whether you can read something 01:14:17.310 --> 01:14:19.680 when you use a proportional display like that. 01:14:19.680 --> 01:14:22.990 And in this case, we're going to have a bunch of big letters, 01:14:22.990 --> 01:14:25.430 and either the letters are going to move slowly 01:14:25.430 --> 01:14:28.030 across the visual field, or your head 01:14:28.030 --> 01:14:30.200 is going to move slowly with the camera like that 01:14:30.200 --> 01:14:32.890 to view that display that's up on the wall. 01:14:32.890 --> 01:14:33.630 OK? 01:14:33.630 --> 01:14:42.540 So if we do that-- are you ready? 01:14:43.860 --> 01:14:45.126 See if you can read it. 01:14:52.680 --> 01:14:53.180 OK. 01:14:53.180 --> 01:14:54.513 Everybody able to read that one? 01:14:55.850 --> 01:14:56.910 OK. 01:14:56.910 --> 01:14:58.704 Who remembers what "fiat lux" means? 01:15:02.620 --> 01:15:04.680 I've only shown it about four or five times. 01:15:04.680 --> 01:15:05.810 Let there be light. 01:15:06.821 --> 01:15:07.320 OK. 01:15:07.320 --> 01:15:12.390 So now, just to be [INAUDIBLE] attractive about it, 01:15:12.390 --> 01:15:15.170 I'm going to show you one more display. 01:15:16.190 --> 01:15:21.670 And this one is a display, again, 01:15:21.670 --> 01:15:27.720 a series of letters and words, is one 01:15:27.720 --> 01:15:40.990 that John F. Kennedy has used in a very, very famous statement, 01:15:40.990 --> 01:15:43.880 which I think most of you will remember. 01:15:43.880 --> 01:15:45.690 And eventually, as you see the whole thing, 01:15:45.690 --> 01:15:47.100 you will be able to put it together. 01:15:47.100 --> 01:15:47.683 Are you ready? 01:17:13.281 --> 01:17:13.780 OK. 01:17:13.780 --> 01:17:15.190 Everybody's seen that, right? 01:17:19.150 --> 01:17:21.310 People nowadays, because of what's going on, 01:17:21.310 --> 01:17:24.370 are less inclined to take his advice. 01:17:33.560 --> 01:17:34.060 All right. 01:17:34.060 --> 01:17:36.680 So that is then the essence of what 01:17:36.680 --> 01:17:38.135 I was going to cover today. 01:17:39.140 --> 01:17:43.610 And so let me now provide you with a brief summary. 01:17:45.430 --> 01:17:48.490 I will start with the prosthetics then. 01:17:48.490 --> 01:17:49.210 OK? 01:17:49.210 --> 01:17:52.410 Research on visual prosthetics is in its infancy, 01:17:52.410 --> 01:17:53.650 as I've noted. 01:17:53.650 --> 01:17:55.880 A great deal of basic research is 01:17:55.880 --> 01:17:58.621 needed before such a device can become effective. 01:17:58.621 --> 01:18:00.120 That's very unlike what you're going 01:18:00.120 --> 01:18:03.620 to hear about the auditory system. 01:18:03.620 --> 01:18:06.660 The brain area that holds considerable promise 01:18:06.660 --> 01:18:11.350 for the prosthetic device based on electrical stimulation 01:18:11.350 --> 01:18:13.340 is area V1. 01:18:13.340 --> 01:18:16.470 Now, a prosthetic device for electrical stimulation of V1 01:18:16.470 --> 01:18:19.270 must take into account the magnification factor. 01:18:19.270 --> 01:18:21.350 That's absolutely essential, and too many 01:18:21.350 --> 01:18:22.890 people are ignoring it. 01:18:22.890 --> 01:18:24.962 Now, in a different color, we're going 01:18:24.962 --> 01:18:26.810 to talk about the illusions. 01:18:26.810 --> 01:18:29.290 There is no unitary explanation for 01:18:29.290 --> 01:18:32.360 the great many visual illusions extant. 01:18:32.360 --> 01:18:34.290 I pointed out to you that different illusions 01:18:34.290 --> 01:18:37.540 have different sources and different explanations. 01:18:37.540 --> 01:18:40.120 The most popular theory explaining the Hermann grid 01:18:40.120 --> 01:18:42.700 illusion based on the center/surround organization 01:18:42.700 --> 01:18:45.650 of retinal ganglion cells is incorrect. 01:18:47.110 --> 01:18:50.050 That's a polite way of saying it's all wrong. 01:18:50.050 --> 01:18:53.630 A more likely theory is the one that 01:18:53.630 --> 01:18:57.285 assumes that area V1 cells are involved. 01:18:58.330 --> 01:19:01.380 Retinal adaptation processes can explain illusions 01:19:01.380 --> 01:19:03.580 based on aftereffects, and you've 01:19:03.580 --> 01:19:05.320 seen quite a few of those. 01:19:05.320 --> 01:19:09.700 And many illusions disappear under isoluminant conditions. 01:19:09.700 --> 01:19:13.830 And lastly, there are no viable theories that explain illusions 01:19:13.830 --> 01:19:16.670 based on figure/ground relationships. 01:19:16.670 --> 01:19:18.440 We know what they are, but we don't 01:19:18.440 --> 01:19:23.350 know how the brain resolves figure and ground effects. 01:19:23.350 --> 01:19:28.120 So that then is the end of what I wanted to cover today. 01:19:28.120 --> 01:19:32.525 And next time, we are going to talk about eye movement. 01:19:33.780 --> 01:19:38.100 And I hope that you will enjoy that as well.