Kristian MyramandClaude Sonnet 5.5 81faf5b4b5 S2.66: mask prompts match Sam2VideoModel ...

Antialiased bilinear resize as PyTorch does it, the reference's mask-as-output
path (mask is the frame's output, the decoder makes the object pointer),
VideoTracker::prompt_mask, and mask-only prompts through the tracker interface.
Prompt frame logits exact, tracking forwards and backwards within 0.0019.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

2026-10-30 08:43:31 +01:00

..

requirements-sam2-reference.txt

S2.0: record the SAM 2 decisions D1-D6

2026-10-30 08:43:29 +01:00

sam2_reference.py

S2.2: make the reference script run without torchvision

2026-10-30 08:43:30 +01:00

sam2_video_reference.py

S2.66: mask prompts match Sam2VideoModel

2026-10-30 08:43:31 +01:00