search

dk-liang / CLTR

[ECCV 2022] An End-to-End Transformer Model for Crowd Localization
MIT License
88 stars 13 forks source link

Test the model wtihout ground truth #6

Closed AhmadrezaAhmadi closed 1 year ago

AhmadrezaAhmadi commented 1 year ago

Hi, Thanks for sharing your code. I was wondering if you have a test code that works with the images without ground truth? In your test.py it seems we need to have shape of kpoint in order to compute topk_values but what if we do not have the ground truth?

dk-liang commented 1 year ago

Yes. You can see the video demo