- Import a video, capture images.
- Train a Dinomaly model.
- Observe the inferenced frames.
The heatmap is shown centrally over the detected object only in the middle of the frame (see attached video). I think it is the matter of scaling. The heatmap images is bigger then displayed frame.
https://github.com/user-attachments/assets/0c1de75c-3ca0-4b0a-924b-0cc6588faa26
The heatmap is shown centrally over the detected object only in the middle of the frame (see attached video). I think it is the matter of scaling. The heatmap images is bigger then displayed frame.
https://github.com/user-attachments/assets/0c1de75c-3ca0-4b0a-924b-0cc6588faa26