search

microsoft / CSWin-Transformer

CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped, CVPR 2022
MIT License
539 stars 78 forks source link

Inference Strategy #38

Open rayleizhu opened 2 years ago

rayleizhu commented 2 years ago
  1. I notice that there is a "--tta" option in args. Do you use TTA in inference to derive the results reported in the paper?
  2. I notice that you specified tricky "model-ema-decay" (e.g. 0.99984 for CSwin-tiny and 0.99992 for CSwin-base), do you use ema-model for inference to derive the results in the paper?
  3. How do the two factors mentioned above impact the model performance?