4 ms·
The chicken and egg problem is that we design neural nets in order for Gabor filters to show up. If you used a different neural net architecture choice, they wo
by argonaut 10y ago
The chicken and egg problem is that we design neural nets in order for Gabor filters to show up. If you used a different neural net architecture choice, they wouldn't show up. So the presence of Gabor filters indicating optimality is sort of begging the question.
- trhway 10y agoCNN beat other architectures in images and Gabor emerges in CNN. That is a pretty good candidate for even global optimaliy.
- argonaut 10y agoThe fact that X works and A, B, and C don't work, does not imply A is globally optimal.
- hacker42 10y agoAs far as I know, Gabor filters show up in all kinds of multi-layered models, even in unsupervised ones. Gabor filters are quite canonical.