Open aashishrai3799 opened 7 months ago
Hi, thanks for your interest. I remember I used the original code from TSM repo to extract video features. I set the base_model='resnet50' , pretrain='imagenet', and num_segments=22 to match the frame numbers in my case. I would probably upload a script later (sorry I don't have it right away). Hope this could help.
@sukun1045 Hi, could you upload the script? That's greatly helpful, thanks!
Hi, thanks for your amazing work! I'm finding difficulty in understanding which part of the code is used to extract video features (before pooling) using TSM? Could you please upload the script to get video features or point me to the file that does that? Your help will be greatly appreciated. Thanks!