10 ms·

Cool idea, but keyframes are not videos. Motion, object permanence, are not things Claude can infer from a set of images. Nice demo though!

• fzysingularity 3mo ago
  Exactly! We experimented with a whole bunch of video encoding techniques for LLMs here: https://vlm-run.github.io/mm/encoders/#video https://vlm-run.github.io/mm/encoders/#video
  • cortexosmain 2mo ago
    [flagged]
• sawjet 3mo ago
  I have been going through this with claude and qwenvl3:8b this week. Both are pretty decent at inferring context and analyzing contact sheets. Finding high visual interest moments with a mixture of coarse and fine keyframes.
  • octember 3mo ago
    Might be time to check gemma :)
• cortexosmain 3mo ago
  [flagged]
• cortexosmain 3mo ago
  [flagged]
• cortexosmain 2mo ago
  [flagged]