DEDUPED Hot AI repos, turned into use cases Get 5 a day

How to check a retargeted clip against its source video with Mixamo LLM Mocap

Render a side-by-side review and measure limb, head and foot errors against the plate.

  • Content
  • Intermediate
  • 1 hour
  • From the README
  • Any agent

Who it is forAn animator or QA reviewer who has a retargeted clip and must decide whether it matches the source video before shipping it.

The problemEyeballing a retarget misses frames where hands, feet or the head drift from the performer, and those errors reach the final clip.

The outcomeA 2x2 review video and a numeric fidelity report that points to the mismatches.

What you need first

  • Blender 5.1 or newer
  • A Mixamo character downloaded as FBX Binary in T-pose from mixamo.com
  • GVHMR cloned into tools/GVHMR with its checkpoints (about 5 GB), set up per docs/INSTALL.md
  • SMPL-X v1.1 body model (free research registration at smpl-x.is.tue.mpg.de), placed as SMPLX_NEUTRAL.npz
  • A GPU with about 8 GB VRAM (developed on an RTX 4080)
  • A locked-camera video plate (filmed or AI-generated)

Steps

  1. Render the review video: python pipeline\render_review.py --spec action_specs\<clip>.json --clip clips\<clip> (inferred, check the README)
  2. Measure the fidelity: python pipeline\eval_fidelity.py --spec action_specs\<clip>.json --clip clips\<clip> (inferred, check the README)
  3. Open the review video and check the plate, overlay, before and after panels, plus the zoomed hands and feet.
  4. If the report flags a window, adjust the action spec (for example fists or arm_overrides) and rerun the retarget and bl_motion.py apply steps.

Prompt for your agent

In the squall01337/mixamo-llm-mocap repo, run render_review.py and eval_fidelity.py on the clip at <clip folder> using its spec <spec path>. Expected output: a review video and a fidelity report with limb and head angles, foot slide and jitter. Verify the report lists foot slide and jitter values for the clip before you call it done.

Caveat. The checks compare against the video's keypoints, so a poor estimate from GVHMR will show up as error rather than being fixed by the review.

Share: X LinkedIn Hacker News Email Copy link

Disclaimer

Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page