# 7.1 AI motion analysis system

## LILLIUS is a specialized exercise service powered by AI human motion analysis technology

The AI motion analysis system is an advanced technology used by LILLIUS to assess the precision of training movements. It is powered by LILLIUS' AI motion analysis technology, which has been developed through extensive research in 2D and 3D camera-based Human Motion Detection since 2016. This innovative system analyzes image data captured by 2D or 3D cameras, evaluates human motions in real-time, including accuracy, timing, and speed, without requiring any additional sensors to be attached to the user's body.

## LILLIUS offers instantaneous analysis and assessment of user actions in real-time

LILLIUS' AI human body motion analysis system compares the user's motion input captured through the camera with expert motions, generating comparative analysis results on the frequency and accuracy of the user's movements. This technology excels in providing evaluations and diagnoses based on real-time accuracy matching, surpassing mere imitation of general videos.

**1. Core posture and motion convergence sensing**

LILLIUS utilizes the smartphone's camera to autonomously extract sports players' motions and create a skeleton exercise database. The process involves human extraction, joint tracking, and bone extraction information stabilization to convert motion details into data. A crucial technology employed by LILLIUS is bone extraction information stabilization, which refines features through key joint selection, creating a hybrid approach that combines key postures' shape and key motions' dynamics. Please refer to section 11 for more details.

**2. LILLIUS ensures the acquisition of accurate exercise movements by providing instant training analysis**

Within the LILLIUS APP, users can utilize their smartphones to analyze and diagnose their workouts while following the training and challenge videos. Notably, LILLIUS' technology allows for the diagnosis of each user's exercise performance, providing immediate real-time feedback. This enables users to correct and learn accurate exercise motions efficiently, as they can apply and repeat the feedback immediately to their subsequent movements.

<figure><img src="https://2827612677-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FEbCQK96f5mOHfNRYXblt%2Fuploads%2FSEXbwZwxG7ZE2GVJmL9U%2F%E1%84%89%E1%85%B3%E1%84%8F%E1%85%B3%E1%84%85%E1%85%B5%E1%86%AB%E1%84%89%E1%85%A3%E1%86%BA%202023-10-06%20%E1%84%8B%E1%85%A9%E1%84%92%E1%85%AE%2011.20.20.png?alt=media&#x26;token=3235365d-4780-494c-b60b-c6be649eb734" alt=""><figcaption></figcaption></figure>

The AI motion analysis system assesses the precision of users' exercise motions and provides a progress score for "following the exercise." With the LILLIUS APP, the AI motion analysis system continually gathers deep learning data, enhancing the accuracy of data evaluation as the LILLIUS platform expands.


---

# 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://lillius.gitbook.io/lillius-white-paper/english/7.-business-and-technical-logics/7.1-ai-motion-analysis-system.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.