About SARM. At this point I am 100% convinced that the VGG16 experiment is not for real. Most likely a big experimental mistake, not fraud.
-
-
Replying to @fchollet
there might be an important technical detail that you missed, eg the "skip connections" or in the way pca/lda is used to get F.
1 reply 0 retweets 0 likes -
Replying to @ogrisel
skip connections are absent in the VGG experiment (0th iteration approximation = relu).
1 reply 0 retweets 0 likes -
Replying to @fchollet
indeed the paper is hard to follow. The sparse coding unrolling story seems decoupled from the experiments.
2 replies 1 retweet 1 like -
I need to read the pcanet paper to understand how lda is used to init conv filters. This should be explained in supplementary mat.
2 replies 0 retweets 0 likes -
Replying to @ogrisel
PCANet does it differently. PCANet was unconvincing, but had accordingly poor results.
2 replies 0 retweets 0 likes
which is great, I should mention. We need more papers that try unconventional ideas, obtain poor results, and talk about it.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.