4 ms·

From my layman point of view, this can lead to two different conclusions: * The DNN is actually getting close to the way human brain works, as it falls into th

From my layman point of view, this can lead to two different conclusions: * The DNN is actually getting close to the way human brain works, as it falls into the same mistakes as the human eye; * There is some kind of universality in these illusions, that every (or most of) vision systems will fall for. But honestly I think that the second conclusion is a rewriting of the first one (what is a vision system? What is this universal property?). Again I'm absolutely not an expert of the field so please correct me if I'm wrong.

candiodari 8y ago
Generally people say that the vision system is a "CNN", convolutional neural net. Now technically speaking a DNN is usually a 5 CNNs in series + another 50 or so "hidden layers" (or 500 if you're Google, or 10000 if you're the human brain), so it's somewhat ambiguous. But that's a "usual" thing. Deep neural nets used to be (20 years ago) anything with more than 1 hidden layer. These days, people say it's anything 50+ hidden layers, because nobody really uses 1 hidden layer for anything but education. These networks types differ in what they compare to what. CNNs compare signals with their neighbourhood (e.g. the pixels in a 9x9 grid), and ignores the rest. DNNs compare every signal with every other signal. CNNs are much cheaper, especially now with dedicated hardware. Large DNNs cannot be accelerated much, and you cannot really build hardware to do that. Also what distinguishes the optical subsystem of animal brains is the first stages are CNNs (well, it's a biological system, so the functions are spread around: the optical sensors in the eye themselves form a CNN, before even sending what they're seeing on, then there's some 5 layers of CNNs in the retina itself before it even gets to the optical nerve, which itself is also a CNN, and delivers the output of that into yet another set of CNNs in your brain).

candiodari 8y ago

Generally people say that the vision system is a "CNN", convolutional neural net. Now technically speaking a DNN is usually a 5 CNNs in series + another 50 or so "hidden layers" (or 500 if you're Google, or 10000 if you're the human brain), so it's somewhat ambiguous. But that's a "usual" thing. Deep neural nets used to be (20 years ago) anything with more than 1 hidden layer. These days, people say it's anything 50+ hidden layers, because nobody really uses 1 hidden layer for anything but education. These networks types differ in what they compare to what. CNNs compare signals with their neighbourhood (e.g. the pixels in a 9x9 grid), and ignores the rest. DNNs compare every signal with every other signal. CNNs are much cheaper, especially now with dedicated hardware. Large DNNs cannot be accelerated much, and you cannot really build hardware to do that. Also what distinguishes the optical subsystem of animal brains is the first stages are CNNs (well, it's a biological system, so the functions are spread around: the optical sensors in the eye themselves form a CNN, before even sending what they're seeing on, then there's some 5 layers of CNNs in the retina itself before it even gets to the optical nerve, which itself is also a CNN, and delivers the output of that into yet another set of CNNs in your brain).