3 ms·

Hm, this seems vaguely interesting-if-true but: 1) It's not published in a good CV journal. 2) It seems quite likely that they've just identified the problems

by nmca 8y ago

Hm, this seems vaguely interesting-if-true but: 1) It's not published in a good CV journal. 2) It seems quite likely that they've just identified the problems described and fixed by [0, 1] (of which [0] is the much better paper). In brief, CNNs guarantee translation invariance (ignoring boundary effects) through translation equivariance of their feature maps; this borks certain problems, most importantly a commonly used mechanism for instance segmentation as described in [0]. Consider trying to instance-segment two copy/pasted faces; translation equivariance and global pooling clearly make it impossible. I strongly suspect that the problem described in the paper originates from a similar place, and thus could be solved in a week or so. I doubt it's a "key weakness of modern cv systems". [0] https://arxiv.org/abs/1807.10712 https://arxiv.org/abs/1807.10712 [1] https://arxiv.org/abs/1807.03247 https://arxiv.org/abs/1807.03247 EDIT - a little further investigation shows that it was accepted as a workshop paper at ICLR, also provides the reviewers comments. https://openreview.net/forum?id=HymuJz-A- https://openreview.net/forum?id=HymuJz-A- So perhaps I'm being overly negative. Can't find any code release for it, would be interested if anyone can. EDIT2 - Entertainingly, the workshop paper version of this and the CoordConv paper seem to use the same name for different toy tasks, "Not-so-CLEVR".

mannykannot 8y ago
Thanks for the link to the reviewers' comments. I am not an academic, and I found some of the objections surprising, such as this one that might be seen as suggesting that the authors just had the wrong attitude: "However, I think there's some chance that if the same tasks were in the hands of people who wanted CNNs or CNN+RN to work well, the results might have been different." The reviewer goes on to write that he cannot say what people with the proper attitude would do differently.
- ColanR 8y ago
  > if the same tasks were in the hands of people who wanted CNNs or CNN+RN to work well, the results might have been different. I guess that's an undesirable symptom of CNNs (etc.) still being an empirical rather than theoretical science.

by nmca 8y ago

mannykannot 8y ago

Thanks for the link to the reviewers' comments. I am not an academic, and I found some of the objections surprising, such as this one that might be seen as suggesting that the authors just had the wrong attitude: "However, I think there's some chance that if the same tasks were in the hands of people who wanted CNNs or CNN+RN to work well, the results might have been different." The reviewer goes on to write that he cannot say what people with the proper attitude would do differently.

ColanR 8y ago
> if the same tasks were in the hands of people who wanted CNNs or CNN+RN to work well, the results might have been different. I guess that's an undesirable symptom of CNNs (etc.) still being an empirical rather than theoretical science.