lzu-siat / as-cal Goto Github PK

Hello,

I ran the code without modification, but unable to reach the accuracy reported in the paper. The reported result on ntu dataset with cross-subject setting is 58%, I ran the original code and only reached 49%. After I normalized the output feature vectors, the accuracy rose to 52%. Could you explain why there is such a huge gap?

Would be appreciated if you could answer this question.
Looking forward.
Thank you.

Hello,
I am using your code, but have some confusing with mom_flag. Would you please answer the following query.

1. You have used three different flags in the code: moco, ins, end, so which one is the main part, and should reflect the reported results in your paper.
2. The train_moco function uses two split, x1 for model training and x2 for model_ema, while the 'ins' used whole data for model training. Even in linear evaluation the ins method is not implemented.

Would be appreciated if you could answer these queries.
Looking forward.
Thank you.