search

Separius / SCALE

official pytorch implementation of "Spatio-Temporal Crop Aggregation for Video Representation Learning" (ICCV23)
4 stars 0 forks source link

Lack some necessary files #1

Open pilibb0712 opened 2 months ago

pilibb0712 commented 2 months ago

Thank you for your awesome work.

I am interested in your work and trying to reproduce it now. However, I have found that there lacks some necessary file. For example, ibot file that is required in src/models/svt.py by 'from models.ibot import DINOHead'. Moreover, some checkpoints do not have download urls like vmae_base1600_pt_kinetics.pth that is required in src/models/interface.py.

Could you please supplement codes and weights that are necessary for reproducing this work.

Thank you very much!

Separius commented 2 months ago

Hi, Thanks for pointing the issues out, I uploaded the missing files. as for the model, I used VideoMAE's checkpoints, I just renamed them.