Transform the caffe model to pytorch model for Temporal Segment Network
The model dir saves my transferred model. I test them on UCF101 and HMDB51 dataset (split1), and get comparable results with the paper as follows.
Dataset | RGB | Flow | Fusion |
---|---|---|---|
UCF101 | 86.01% | 87.70% | 93.82% |
HMDB51 | 54.90% | 63.53% | 71.18% |