search

broadinstitute / keras-resnet

Keras package for deep residual networks
Other
300 stars 127 forks source link

Difference with 0.2.0 vs 0.1.0 #57

Open hgaiser opened 5 years ago

hgaiser commented 5 years ago

The network architecture in 0.2.0 is no longer identical to that in the original Caffe implementation. I am using the ResNet50 imagenet weights available here, which were generated using the caffe weights and this tool.

On some Tabby Cat image I get the following predictions from 0.1.0:

[[('n02123045', 'tabby', 0.4307788), ('n02124075', 'Egyptian_cat', 0.32408533), ('n02123159', 'tiger_cat', 0.18477823), ('n02127052', 'lynx', 0.008598777), ('n03443371', 'goblet', 0.0048426227)]]

On 0.2.0:

[[('n02123045', 'tabby', 0.39859685), ('n02124075', 'Egyptian_cat', 0.26450825), ('n02123159', 'tiger_cat', 0.22655205), ('n02127052', 'lynx', 0.012361869), ('n03443371', 'goblet', 0.00793589)]]

If I find some more time, I will also show the results using Caffe. But for now, it is clear that there is an unintended difference between 0.2.0 and 0.1.0. I believe this difference comes from https://github.com/broadinstitute/keras-resnet/commit/5005c374fd58b9cea53c0fa640ddb159e0f4bb0f . @0x00b1 do you know why that commit was applied?

hgaiser commented 5 years ago

@0x00b1 , any feedback on this?