search

implus / PytorchInsight

a pytorch lib with state-of-the-art architectures, pretrained models and real-time updated results
858 stars 123 forks source link

about group conv #5

Open valencebond opened 5 years ago

valencebond commented 5 years ago

As paper said，group conv of different group channel can extract different part features of an object.I wonder the improvement of performance come from group conv or attention mudule. Do you run an experiment of setting group num == 1 in Fig5? if you can compare vanilla resnet50 perf and resnet50 with group conv perf，it may be more convinced.

implus commented 5 years ago

Thanks for suggestions~ We will add these experiments in the extended version.