# How does it work ?

<figure><img src="/files/jLuCZScLhXk8jzoZ8EO2" alt=""><figcaption></figcaption></figure>

PoseTracker exposes a simple URL that runs pose estimation and motion tracking.

It now uses a **single tracking engine (V3)** for both real-time and upload flows.

### Client-side by design

* ❌ no streaming
* ❌ no latency
* ❌ privacy problems

### What you get

* Real-time pose estimation (17 keypoints).
* Optional skeleton overlay.
* Optional exercise tracking (posture, counter, progression, recommendations).
* Optional raw outputs (keypoints, angles) on supported plans.

### Model

PoseTracker is based on TensorFlow MoveNet by default.

You can also enable BlazePose (plan-gated).

### Typical flow (V3)

1. Initialization.
2. Placement guidance.
3. Optional countdown.
4. Active tracking:
   * posture + counter
   * optional form score, progression, recommendations
   * optional keypoints / angles

For upload flows:

* video ends with a final counter
* export finishes with `export_ready` when enabled

It can run at high FPS on modern devices.

### Pose estimation

<figure><img src="/files/oG4TkQ8BXAJyBIplNzJp" alt=""><figcaption><p>PoseTracker keypoints</p></figcaption></figure>

### Motion analysis

PoseTracker analyzes keypoints over time.

This enables rep counting and movement feedback.

### Try it

{% hint style="info" %}
The easiest way to experience PoseTracker is the mobile demo app.

Install it on iOS or Android.

Then follow along with the docs.

* iOS (App Store): <https://apps.apple.com/us/app/posetracker-ai/id6759670702>
* Android (Google Play): <https://play.google.com/store/apps/details?id=com.posetracker.v3demo>
  {% endhint %}

Prefer a browser-based sandbox?

See [Try it yourself](/posetracker-api/use-posetracker-on-real-time-camera-webcam/try-it-yourself.md).


---

# Agent Instructions: 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:

```
GET https://posetracker.gitbook.io/posetracker-api/use-posetracker-on-real-time-camera-webcam/how-does-it-work.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.