tao-generate-video-reasoning-annotations
SkillAbout
Multi-step video annotation pipeline that turns raw videos into Chain-of-Thought training data — multi-level captions, structured descriptions, and QA pairs (MCQ, binary, open-ended) with reasoning traces, via VLM/LLM distillation. Use when the user wants to "create video training data", "generate video QA datasets", "build CoT reasoning traces from videos", "auto-label videos", or run the video_reasoning_annotation pipeline. Triggers include "video annotation", "video CoT", "video QA", "chain-of-thought", "video captioning pipeline", "video distillation".
Capabilities
The crawler did not record capability metadata for this resource. Inspect the endpoint
directly to see what it exposes.
Provenance
Discovered Relayed by agntcy
URN authority urn:air:outshift.io:agntcy:tao-generate-video-reasoning-annotations
Catalog host outshift.io
Anchor check Not anchored
Last crawled seen 5h ago