search

facebookresearch / DetectAndTrack

The implementation of an algorithm presented in the CVPR18 paper: "Detect-and-Track: Efficient Pose Estimation in Videos"
Apache License 2.0
1k stars 186 forks source link

about the test #46

Closed ezreal1129 closed 6 years ago

ezreal1129 commented 6 years ago
  1. My data is only ground truth, no keypoints.so can i test the pre-model?
  2. The key point of the data results is how to form the boundingbox data, because I can only use the boundingbox to evaluate.
rohitgirdhar commented 6 years ago

Yes, please try tools/test_on_single_video.py to test pretrained models, which doesn't need any GT annotation.

ezreal1129 commented 6 years ago

@rohitgirdhar Thank you ,but how should i evaluate the rusult of test_on_single_video.py