2025-06-24 14:20:02 +08:00
# MNN-TaoAvatar Android - Local 3D Avatar Intelligence
2025-03-21 15:00:13 +08:00
2025-06-24 14:38:55 +08:00
< div align = "center" >
< b style = "color:red;" > ⚠️ Attention: this folder has been moved to: < / b >
< a href = "../MnnTaoAvatar/README.md" > apps/Android/MnnTaoAvatar< / a >
< / div >
2025-06-11 11:47:18 +08:00
< p align = "center" >
< img width = "20%" alt = "Icon" src = "https://meta.alicdn.com/data/mnn/avatar/avatar_demo.gif" style = "margin: 0 10px;" >
< / p >
2025-06-24 14:27:55 +08:00
+ [中文 ](../MnnTaoAvatar/README_CN.md )
+ [Download ](../MnnTaoAvatar/README.md#releases )
2025-06-11 11:47:18 +08:00
2025-06-24 16:15:22 +08:00
Welcome to **MNN-TaoAvatar App** , a full-featured mobile app about the [TaoAvatar paper (arXiv:2503.17032v1) ](https://arxiv.org/html/2503.17032v1 ). This project brings multimodal AI avatars to life directly on Android devices, running **all models locally** , including:
2025-06-11 11:47:18 +08:00
- LLM (Large Language Model)
- ASR (Automatic Speech Recognition)
- TTS (Text-to-Speech)
2025-06-19 16:11:49 +08:00
- A2BS (Audio-to-BlendShape)
2025-06-11 11:47:18 +08:00
- NNR (Neural Rendering)
> The iOS App will be coming later, stay tuned for updates!
2025-06-11 12:13:06 +08:00
## Features
2025-06-11 11:47:18 +08:00
- **Conversational AI** powered by a local LLM
- **Speech-to-text** with embedded ASR models
- **Voice synthesis** with TTS on-device
- **Avatar behavior animation** via A2BS(Audio-to-BlendShape)
- **Real-time neural rendering** for expressive avatars
- **100% offline** and privacy-focused
2025-06-11 12:00:13 +08:00
## How it Works
2025-06-24 14:27:55 +08:00

2025-06-11 11:47:18 +08:00
2025-06-11 12:13:06 +08:00
## Requirements
2025-06-11 11:47:18 +08:00
Because **all AI models are executed locally on-device** , this project requires **high-performance hardware** to run smoothly.
### Minimum Device Requirements
- **Snapdragon 8 Gen 3** or **equivalent flagship SoC**
_Examples: Snapdragon 8 Gen 3, Dimensity 9200 to have smooth experience._
- **8 GB RAM** or more
- **5 GB free disk space** for model files
- **ARM64 architecture**
> ⚠️ Devices below these specs may experience lag, audio stutter, or limited functionality.
2025-06-11 12:13:06 +08:00
## Setup and Run
2025-06-11 11:47:18 +08:00
### 1. Clone the Repository
```bash
git clone https://github.com/alibaba/MNN.git
2025-06-24 14:20:02 +08:00
cd apps/Android/MnnTaoAvatar
2025-06-11 11:47:18 +08:00
```
### 2. Build and Deploy
- Connect an Android device
- Hit **Run** in Android Studio or use:
```bash
./gradlew installDebug
```
## Links And References
- [TaoAvatar Paper ](https://arxiv.org/html/2503.17032v1 )
- [Models Collection ](https://modelscope.cn/collections/TaoAvatar-68d8a46f2e554a )
- [LLM Model: Qwen2.5-1.5B MNN ](https://github.com/alibaba/MNN/tree/master/3rd_party/NNR )
- [TTS Model: bert-vits2-MNN ](https://modelscope.cn/models/MNN/bert-vits2-MNN )
- [A2BS Model: UniTalker-MNN ](https://modelscope.cn/models/MNN/UniTalker-MNN )
- [NNR Model: TaoAvatar-NNR-MNN ](https://modelscope.cn/models/MNN/TaoAvatar-NNR-MNN )
- [ASR Model: sherpa-mnn-streaming-zipformer-bilingual-zh-en-2023-02-20 ](https://modelscope.cn/models/MNN/sherpa-mnn-streaming-zipformer-bilingual-zh-en-2023-02-20 )