Skye
Skye
> @17Skye17 It seems that the author will remove '-' and '_' chars when a video id starts with these two. > > And @sallymmx Could you please provide the...
@achaiah Thank you for comment. I built a new training framework based on this repo :-)
I have the same request.Thank you!
Thank you for reply. @GV1028 I got what you mean, I just think line 35 is werid, I guess the original video frames are in range [0,1],so they should be...
Same doubts as yours. @avisekiit
Thank you!I'll try it @ahmed-fau
Hi, @patykov @Tushar-N, sorry for late reply, I've been busy the last two weeks. Thank you for your comments, I'll check my results again and find out what is wrong...
Hi, this codebase does not provide a dataloader to read frames from raw videos or feature extractor, you can refer to this [repo](https://github.com/m-bain/frozen-in-time/blob/e1f13e178ce4d7ae5ef8a54232df46b5603194bf/base/base_dataset.py#L15) to read raw videos and this [repo](https://github.com/mit-han-lab/temporal-shift-module)...
I'm just a passerby, for loading large dataset, maybe this website will help. @lkffight Loading a HDF5 file is much faster than loading images from folders. : ) http://machinelearninguru.com/deep_learning/data_preparation/hdf5/hdf5.html