domains / families / media / rn-human-motion-extractor

rn-human-motion-extractor

separate

Extract anonymized frame-by-frame body and hand trajectories from an authorized reference video, with skeleton previews, confidence evidence, and machine-readable keypoints. Use when the user asks to 提取博主动作, 提取人体动作轨迹, 做姿态参考, 生成骨架视频, or prepare motion control data for an avatar; do not route graphic-animation replication here.

When to use

• extracting body and hand trajectories
• generating skeleton videos
• preparing motion control data for an avatar
• creating pose references
• extracting human motion trajectories

When not to use

• graphic-animation replication
• designed cards, connectors, typography, and UI motion
• full-frame replica claims
• packaging source person's appearance, voice, subtitles, or footage as a reusable identity asset

Skill page

Source