Virtual live-streaming assistant for platforms like Bilibili
Top 91.7% on sourcepulse
This project provides a Python-based virtual streamer assistant that integrates with Bilibili, Douyin, and WeChat Channels to capture live stream comments. It leverages ChatGPT for natural language understanding and response generation, and Edge TTS for voice synthesis, aiming to enhance viewer interaction and immersion for VTubers.
How It Works
The system employs a modular architecture. A "Danmaku Acquisition Module" uses blivedm
and DouyinBarrageGrab
to fetch comments from various platforms. These comments are then processed by a "Semantic Understanding Module" powered by ChatGPT for intelligent replies. A "Voice Synthesis Module" utilizes Edge TTS to convert text replies into speech, which is then fed to the virtual avatar. Finally, a "Screen Push Module" integrates with OBS to display the virtual host's speaking animation in the streaming software.
Quick Start & Requirements
pip install -r requirements.txt
config.py
with Bilibili room_id
and openai.api_key
.douyin.py
with openai.api_key
.chromedriver
, scan QR code for login cookies, then run wechat.py
.python bilibili.py
WssBarrageService.exe
then python douyin.py
Highlighted Details
Maintenance & Community
The project relies on external open-source libraries blivedm
and DouyinBarrageGrab
. Further details and support are available via a "Knowledge Planet" mentioned in the README, which reportedly includes API key sharing and ChatGPT tutorials.
Licensing & Compatibility
The project is intended for learning and research purposes only and explicitly prohibits commercial use. Users are advised to comply with the terms of service for Bilibili, Douyin, and other platforms.
Limitations & Caveats
The project is strictly for non-commercial use. Users must adhere to platform usage policies. Detailed tutorials and API key sharing are available through a paid "Knowledge Planet" service.
2 years ago
1 week