> For the complete documentation index, see [llms.txt](https://doc.ancorasir.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.ancorasir.com/2026spring/final-project-example-human-gesture-to-humanoid-motion.md).

# Final Project Example: Human Gesture to Humanoid Motion

This worked example demonstrates the technical core of the final project: record a human gesture, recover a whole-body motion representation, and retarget that motion to a humanoid. It is a compact reference rather than a complete submission; student reports should also document the gesture's social context, data protocol, method, quantitative evaluation, limitations, and reproducible code.

### Example pipeline

| Stage             | Example method                                          | Output                                         | Video                                                                                                                  |
| ----------------- | ------------------------------------------------------- | ---------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------- |
| 1. Record         | RGB video of a participant raising the arms in sequence | Source gesture video                           | Local video · [GitHub-hosted preview](https://github.com/user-attachments/assets/f6b454f6-721a-4c36-9436-6c9782b6bb3b) |
| 2. Recover motion | GVHMR (Global-View Human Motion Recovery)               | SMPL-X motion data and body-pose visualization | Local video · [GitHub-hosted preview](https://github.com/user-attachments/assets/1a701bdd-fb81-4f59-926e-daf4af9316a7) |
| 3. Retarget       | General Motion Retargeting (GMR)                        | Unitree G1 joint motion                        | Local video · [GitHub-hosted preview](https://github.com/user-attachments/assets/b11f90ba-ff31-4ae7-a73b-286efaa911eb) |

### 1. Record the source gesture

Tianyu demonstrates a whole-body gesture in which the arms are raised in sequence. A full project should describe the recording device, camera placement, frame rate, resolution, participant position, number of takes, and any privacy processing. Record enough context to make the gesture's timing and intended meaning clear.

### 2. Recover the human motion

The example uses GVHMR to recover body motion from the RGB video and stores the result in an SMPL-X-compatible representation. The recovered data is visualized before retargeting so pose errors, temporal jitter, global drift, and missing observations can be identified early.

A full submission should state the exact model version and output schema, explain coordinate frames and units, and report at least one quantitative measure of reconstruction quality or temporal consistency.

### 3. Retarget to a humanoid

The recovered SMPL-X motion is retargeted to Unitree G1 with GMR. The final video should be evaluated against the source motion. Useful measures include key-body position and orientation error, joint-limit violations, stance-foot slip, self-collision count, motion smoothness, and timing alignment.

Other motion-recovery, robot, or retargeting methods are welcome. The important requirement is a justified and reproducible pipeline with evidence that each stage works.

### Suggested submission checklist

* [ ] Explain the gesture and its positive social meaning.
* [ ] Document a reproducible recording and data-processing protocol.
* [ ] Describe the recovered motion format, frames, units, and timing.
* [ ] Visualize and evaluate motion recovery before retargeting.
* [ ] Explain the target robot and retargeting algorithm.
* [ ] Compare the retargeted robot motion with the source gesture.
* [ ] Include runnable code and environment/setup instructions.
* [ ] Discuss limitations and next steps toward hardware deployment.

***


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://doc.ancorasir.com/2026spring/final-project-example-human-gesture-to-humanoid-motion.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.