An open-source project dedicated to tracking and segmenting any objects in videos, either automatically or interactively. The primary algorithms utilized include the Segment Anything Model (SAM) for key-frame segmentation and Associating Objects with Transformers (AOT) for efficient tracking and propagation purposes.
I am utilizing your code to segment and track multiple fish in a video. However, I am encountering an issue where the algorithm detects 96 fish instead of the 32 selected after frame 99, including numerous false positives in the background. Could there be any hardcoded parameters causing this error from frame 100 onwards? Additionally, are there any adjustments I can make to reduce these false positives?
I am utilizing your code to segment and track multiple fish in a video. However, I am encountering an issue where the algorithm detects 96 fish instead of the 32 selected after frame 99, including numerous false positives in the background. Could there be any hardcoded parameters causing this error from frame 100 onwards? Additionally, are there any adjustments I can make to reduce these false positives?