search

laura-wang / video-pace

code for our ECCV-2020 paper: Self-supervised Video Representation Learning by Pace Prediction
98 stars 12 forks source link

Can you release your code about contrast learning? #4

Closed huangjinghuangjing closed 3 years ago

huangjinghuangjing commented 3 years ago

When reading your code, I found that there seems to be a lack of code for contrast learning！

baoxin1100 commented 3 years ago

I found nothing about contrast learning yet!

laura-wang commented 3 years ago

Thanks for your interest. Sorry, I am quite busy these days... The contrastive learning we used is similar as the SimCLR framewrok. I believe it is not difficult to implement by yourself. I will clean and release the code when I am not so busy....Please stay tuned.

RajArPatra commented 3 years ago

@laura-wang any updates on the contrastive learning code?

laura-wang commented 3 years ago

Thanks for your interest. Sorry, I am quite busy these days, working some new topics. I will release the code as soon as possible (hopefully)...

If you are in a hurry to use the contrastive learning code for video, facebook has conducted a large-dcale study "A Large-Scale Study on Unsupervised Spatiotemporal Representation Learning" , and the code is releasedat https://github.com/facebookresearch/SlowFast, please first refer to this code.