Code for "Spatiotemporal Contrastive Video Representation Learning"

[napohou]

Hi all,
How can we perform a downstream task of action detection?

[napohou]

The official code is here: https://github.com/tensorflow/models/tree/master/official/vision/beta/projects/video_ssl
But it looks no evaluation code for action detection.

[napohou]

any update?