The implementation codes on Caffe for ECCV 2018 paper DA-Net.
We use the Temporal Segment Network as our backbone and make multi-branch neural networks for the multi-view action recognition.
The instructions on using the codes will be updated soon.