Mark12Ding / FAME

Code for Motion-aware Contrastive Video Representation Learning via Foreground-background Merging (CVPR 2022)
https://arxiv.org/abs/2109.15130
Apache License 2.0
45 stars 5 forks source link

Code for violin-plotted and CAAM #7

Open fmthoker opened 1 year ago

fmthoker commented 1 year ago

Hi, Thanks for releasing the code of your paper. Is it possible to share the code that you have used to generate Figure 1 and Figure 3? I am currently working on something similar and would like to compare it with FAME on these axes. I am also using the R2+1D-18 network.