AIAny
AI Audio2024
Icon for item

Voice-Pro

Local WebUI for video and audio dubbing: download YouTube, isolate vocals, transcribe with Whisper, translate into 100+ languages, and generate multilingual TTS or zero-shot voice clones. Uses Gradio, yt-dlp, Demucs, Whisper/Faster-Whisper, F5/E2/CosyVoice and Edge-TTS; Windows-focused with optional Azure integration.

Visit Website

Introduction

Most content creators still stitch multiple tools to turn a foreign-language video into localized subtitles and a dubbed track. This project bundles that pipeline into a single local WebUI so you can download a YouTube video, separate vocals, run ASR, translate subtitles, and synthesize a dubbed track without juggling separate CLIs and manual format conversions.

What Sets It Apart
  • Unified, local workflow: integrates yt-dlp, Demucs (vocal isolation), Whisper/Faster-Whisper (ASR), Deep-Translator (100+ languages) and multiple TTS/voice-cloning backends so users work in one Web interface instead of many disconnected tools — so what? faster iteration and fewer format/encoding hassles when producing subtitles and dubbed audio.
  • Zero-shot voice cloning + multiple TTS providers: supports F5-TTS, E2-TTS, CosyVoice (including Fun-CosyVoice3) and Edge-TTS, with optional Azure keys for higher-quality/enterprise endpoints — so what? you can produce voice-matched dubs across many languages without training custom models.
  • Practical, reproducible installer: uses uv lockfiles and portable dependencies (auto-downloads ffmpeg, bundled Python runtime, prebuilt wheels) so installs are fast and reproducible on Windows machines without requiring CUDA toolkit or Visual Studio — so what? reduces setup friction for creators on restricted or corporate PCs.
  • Open-source, Windows-first design with GPU acceleration: optimized for NVIDIA GPUs (works on CPU/GPU), exposes denoise/compute-type controls and subtitle formats (SRT/ASS) for downstream editing.
Who it fits / trade-offs

Great fit if you need an all-in-one, locally run dubbing/subtitle pipeline for podcasts, YouTube content, or multilingual video localization and you prefer an offlineable toolchain. Look elsewhere if you need a hosted SaaS with guaranteed SLAs, a cross-platform polished Mac/Linux UX (the project is Windows-focused and best-tested on NVIDIA GPUs), or if legal/ethical constraints forbid the use of celebrity voice samples — the repo includes many reference voices and relies on user responsibility for voice consent.

Back

Information

  • Websitegithub.com
  • OrganizationsABUS, WeConnect
  • AuthorsABUS
  • Published date2024/07/29

Categories

Tags

More Items

Hugging Face
AI Audio2026
Icon for item

AuK: An Open-Source Foundational Model for Speech Generation and Editing

Ziyang Ma, Zhikang Niu +31Tencent (Tencent Hunyuan)

Generates and edits speech from natural-language instructions plus optional reference audio, supporting zero-shot TTS, content/acoustic/paralinguistic edits, enhancement, and source separation. Open-source 1.5B-parameter base model with a 4-step distilled AuK‑Flash for faster inference.

audiospeechttsvoicefoundation-model+5
Hugging Face
AI Audio2026
Icon for item

YuE2-3B

Ruibin Yuan, Hanfeng Lin +56Multimodal Art Projection (MAP), Hong Kong University of Science and Technology (HKUST)

Generates full songs from lyrics and a style prompt, producing vocals and accompaniment and exporting editable symbolic scores for melody/chord control. Uses symbolic planning with agentic multi-turn editing and runs local 48 kHz inference on a 24GB GPU.

flow-matchingsafetensorspytorchhuggingfacevllm+7
Hugging Face
AI Dataset2026
Icon for item

Text-to-Speech Human Preferences (315K)

Datapoint AI

Provides 315,000 pairwise human-preference votes comparing 15 English TTS models over 300 operational prompts, with 4,500 high‑quality audio renders and structured vote/pair/prompt records for training or evaluating preference/reward models. Metadata under CC-BY-4.0; audio use governed by model providers' terms.

ttsaudiospeechbenchmarkevaluation+4