You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
first of all I want to thank you for your work and making it accessible to everyone.
I want to work with the trained corridor cobevt model, but whenever I run the inference with the provided model checkpoint, the results are quite bad and not close to the reported results. In fact, AP=0 for IoU=0.3, 0.5, and 0.7.
Would you perhaps know what the source of the problem might be?
I'll attach a screen capture of an inference for one frame (red is the output of the model, green the ground truth).
Thank you!
The text was updated successfully, but these errors were encountered:
Hello,
first of all I want to thank you for your work and making it accessible to everyone.
I want to work with the trained corridor cobevt model, but whenever I run the inference with the provided model checkpoint, the results are quite bad and not close to the reported results. In fact, AP=0 for IoU=0.3, 0.5, and 0.7.
Would you perhaps know what the source of the problem might be?
I'll attach a screen capture of an inference for one frame (red is the output of the model, green the ground truth).
Thank you!
The text was updated successfully, but these errors were encountered: