scwf/open-dubbing

Overview

The AI voiceover tool is a professional AI voice cloning solution that uses advanced TTS models to convert SRT subtitle files or TXT text into high-quality dubbing audio. The tool supports multiple time synchronization strategies, accurately matching subtitle durations to generate voiceovers perfectly synchronized with videos. - : Supports time stretching strategies to ensure dubbing and subtitle durations match perfectly - : Based on models like Fish-speech\IndexTTS2\CosyVoice\F5 to generate natural and smooth speech - : Provides basic and stretch strategies to meet different needs - : The IndexTTS2 engine supports emotional expression control, allowing adjustment of voice emotion through audio, vector, text, and other methods - : Offers an intuitive and user-friendly web UI, supports file drag-and-drop upload, online parameter configuration, and real-time progress display, greatly simplifying the operation process. - : Professional log system displays processing progress and status in real time - : Three engines lock upstream commits via Git submodule; , provided by pip install f5-tts. - : Already in .gitignore, not included in the versions; Automatically downloaded to this directory after running the corresponding install-*.sh. - : Platform-independent skill pack for AI Agents (see Agent Skill below); SKILL.md Defines the voiceover workflow, reference.md provides extended reading on engine, parameters, and troubleshooting. - : Web UI (server.py / run.sh) and CLI (run_dubbing.py) share the same src/ logic. This project relies on the following third-party source code repositories and is fixed to a verified commit, not drifting with the latest upstream version: If you encounter a failed download of deps/index-tts Git LFS during cloning (for example, temporary upstream LFS quota exceedance), you can temporarily skip the LFS smudge: - The installation script will only initialize the submodule version locked in the repository and will not automatically pull the latest upstream source code. - If the submodule has not been initialized, please execute the above command first, then run the installation scripts for each engine The project provides run.sh one-click deployment scripts that can automatically configure environments, install dependencies, download models, and start services: - 🔧 Automatically create and activate the index-tts25 Conda environment - 📦 install all required dependencies (including FFmpeg, PyTorch, etc.) - 🔗 initialize and install the locked version of IndexTTS-2.

Overview (machine translated)

This page was translated automatically. Technical details may differ from the original.

Recommended Tools

Try a different keyword or remove a filter.

Install