trwappers
2 HN recommenders · recommended in 4 months, 2017–2018
1:11:36 · 200K views · Apr 3, 2017
Geoffrey Hinton gives an MIT talk on what is wrong with convolutional neural networks and introduces his capsules project. The description points to the capsule network papers.
Recommended as Hinton's own explanation of capsule networks and his critique of pooling in CNNs.
5 comments from Hacker News
Hinton explains the concept of capsules in this video: this video Which is a lot better than reading someone tell you about this new idea that is called "capsules" but doesn't go into detail. The only thing is that, when this presentation was given, it seems they hadn't worked much more than MNIST (so the new thing now would be the toys-recognition net). Better source, with date: [LINK] (December 2014, for the lazy).
Prof. Hinton has an interesting talk about his new 'capsule' model based on psychology of shape perception: this video
I'm certain that AI business will plateau and there will be yet another AI winter without new big breakthroughs in fundamental research. We are currently living in a era where the results from Hinton & Canadian Mafia are still begin adopted and refined. It's just alternating layers of affine transformations and nonlinearity with lots of tricks and improved routing. There needs to be fundamental theoretical breakthrough in every decade or so, and not just refinements and new applications. Hinton made his "What is wrong with convolutional neural nets" speech years ago " this video I believe that Hinton's capsule networks type attempts are what we need more. They may not deliver immediately impressive results because fundamental research involves lots of setbacks.
I have a hypothesis that these errors in interpretation stem mainly from the max pooling layers that occurs in nearly all CNNs today. Simply cutting out a large fraction (at least 75%!) of the information passing through each pooling layer is making the models quite brittle. That, and obviously it would be a big help to be able to interpret meaningful symbols (e.g. words) from the images rather than just recognize a certain arbitrary class of pattern activations. That is pretty hard, of course, but IMO essential for the upcoming self-driving car revolution, if it is to succeed. G. Hinton has some ideas though they aren't very fleshed out yet.[1] [1] this video
Hinton's "capsule networks": [LINK] this video [LINK] They sound a bit like "grandmother neurons/cells": [LINK] I would very much appreciate an ELI5 of capsules and some prognostications about how capsules may affect the state of the art in neural networks and the current flood of papers and graduate students in the pipeline of ML.