01
Upload and trim the source clip
Start with a short performance video and trim the exact motion window you want to process so the workflow focuses on the usable action range.
AI VIDEO MOCAP
Turn one short clip into animation-ready motion and robot motion outputs
AIMoCap combines AI video mocap, markerless motion capture, animation-ready FBX delivery, and robot motion outputs in one browser workflow so teams can move from source video to downloadable results faster.
What this workflow is optimized for
- AI video mocap from one short uploaded clip
- Markerless motion capture without a suit or markers
- Animation-ready FBX output for review and cleanup
- Robot motion collection with Unitree G1 delivery
Who this page is for
AIMoCap is built for animation teams, technical artists, and robotics teams that need motion capture from video without running a suit-and-marker capture stage first.
If your workflow is upload a short clip, review the motion, and then download FBX or robot JSON, this page covers the browser-based video mocap path that AIMoCap is built to support.
Typical workflow
02
Send one clip to multiple targets
Submit the same clip to Default, Unitree G1, and any published custom avatar targets in one run so review and delivery stay aligned.
03
Review before you download
Open the result page to compare processed video and 3D output before downloading FBX or robot motion JSON.
What you can review and deliver
Default FBX motion
Generate animation-ready FBX output for review, cleanup, and downstream DCC workflows.
Unitree G1 robot motion
Send the same clip into a robot motion collection workflow and review Unitree G1-ready motion JSON outputs.
Custom avatar retargeting
Reuse published custom avatars as output targets so one clip can drive Default, Unitree G1, and custom character motion in the same run.
Reviewable processed results
Compare processed video and 3D motion before download so your team can review motion quality instead of guessing from raw files.
Common questions
What kind of source video works best?
Short clips with a clear subject, stable framing, readable body motion, and no hard cuts usually produce the best AI video mocap results.
Is AIMoCap a markerless motion capture workflow?
Yes. AIMoCap is designed as a markerless motion capture workflow that starts from source video instead of a suit-and-marker capture stage.
Can one clip produce both Default and Unitree G1 outputs?
Yes. AIMoCap can send the same clip to Default and Unitree G1 in the same job so you can review both output types together.
Can I use my own character?
Yes. After a character passes the binding and test workflow, it becomes a reusable custom avatar target inside Studio.
What file outputs do I get?
Default targets can deliver FBX plus processed preview video, while Unitree G1 targets can deliver robot motion JSON plus previewable result output.
Is there a video mocap API for automation?
Yes. AIMoCap provides an async video mocap API so teams can upload clips, poll job status, review results, and download outputs inside an existing production pipeline.