bilivid-metaprocess-server

A server that enhances bilibili video processing with advanced features.

Features

• Music recognition in videos via ACRCloud API

More features are under development.

Installation (Docker)

1. Create a directory for the application, e.g., /opt/bilivid-metaprocess-server.
2. Clone the repository into the directory: git clone https://github.com/fred913/bilivid-metaprocess-server.git /opt/bilivid-metaprocess-server/src.
3. Copy docker-compose.yml from the repository to the app directory.
4. Adjust docker-compose.yml as needed. For example, modify port mapping or add environment variables.
5. Input your ACRcloud API credentials into the .env file (refer to .env.example for the format).
6. Launch the application: docker-compose up -d.
7. Due to occasional issues causing the application to panic, it's advised to restart the application every minute. Use a shell script (run it every minute via crontab or similar):

#!/bin/bash
cd /opt/bilivid-metaprocess-server
docker-compose restart

Usage

The server operates over HTTP and offers the following endpoint:

• /api/recog_music_in_bili_video: Accepts three parameters: video_aid: int, pid: int, target_sec: int. Returns the same format as the ACRCloud API.

License

MIT

To-do

• STT Provider: OpenAI Whisper API

• STT Provider: OpenAI Whisper (open-source)

• STT Provider: other cloud providers'

• Recognize music from a video using ACRCloud API

• Speech-to-text recognition from a video using STT

• Enhance music recognition by (optinally) removing vocal from the audio signal (by MDX or similar)

• Pre-built Docker image