Open tuandat-github opened 1 year ago
which layer should I use for feature extraction to get the best results( example: layer2 and layer3 in wide resnet50).
I also found this a problem, ended up searching which layers exist inside the network architecture, but when I tried to pick one, it gives error that the sizes don't match. Any luck figuring this out?
which layer should I use for feature extraction to get the best results( example: layer2 and layer3 in wide resnet50).