search

Ignotus / theano-flownet

FlowNetS and FlowNetC port to Theano
MIT License
8 stars 4 forks source link

Batch Norm #3

Closed pvskand closed 6 years ago

pvskand commented 6 years ago

In the original flownet implementation there is batch normalization after every conv layer. In theFloatNetC.py, I don't find any batch_norm layer. Any specific reason?

Ignotus commented 6 years ago

Which FlowNet configuration are you talking about? I couldn't find any batch normalization in their protobuf file.

pvskand commented 6 years ago

Okay so I was looking at NVIDIA's Pytorch implementation of Flownet 2.0. In that they had used batch_norm so I assumed the case with Flownet 1.0, but it's not the case.