A fast python library for aligning similar audio snippets passed in as NumPy arrays
-
Updated
Oct 27, 2025 - Python
A fast python library for aligning similar audio snippets passed in as NumPy arrays
Verify that a subtitle actually matches and syncs with a video using audio-based timing analysis. Detect wrong subtitles, find the correct timing offset, and automatically sync them.
Estimation of audio alignment relative to a given reference.
Align various Sanskrit texts and audio
Python CLI to facilitate making karaoke videos
Multi-track audio timebase estimation, alignment and clock-drift correction for recordings from independent devices. TimeMap abstraction, GCC-PHAT fine alignment, drift/changepoint estimation, multi-track graph solver, validation, Adobe Audition SESX export, and a dedicated MP4 wireless-microphone channel-sync pipeline.
歌词字幕对齐命令行工具:demucs 起唱点提取 + 带约束单调动态规划,中位偏差 0.00s(人工校验真值对照)
Local-first toolkit for fixed-camera band rehearsal videos: YAMNet song/highlight/fun-moment detection, multicam-style auto editing on virtual cameras, lossless raw-cut export, and a reusable audio-to-video alignment library for swapping in a clean recorder take of the same performance.
Danmaku Studio 弹幕一站式解决方案
MMS Forced alignment package for JavaScript, TypeScript and Node.js: word-level timestamps from audio + transcript for captions, subtitles and karaoke. Meta MMS (wav2vec2) via ONNX Runtime with a CTC Viterbi decoder. Hindi, Hinglish and 20 Indic languages. No Python.
🎧 Align audio to text effortlessly with precise word-level timing using OpenAI Whisper, outputting clean JSON for seamless integration.
Non-destructive WAV/MP3 and TextGrid timeline alignment for Praat 7 on macOS, with gapless decoding and multi-pair opening.
Open toolkit for human-directed AI music-video production: storyboards, animatics, prompts, media review, audio alignment, lip-sync research and GPU workflows.
Align and graft dubbed audio across media releases using distributed audio anchors.
Word-level Bible audio alignment data produced by Snail
RAB (RysUp Align Bench): an audio-to-audio vocal alignment benchmark with real and exact timing truth, reference aligners, and the RysUpAlign learned alignment model v6
複数トラック音源のサンプリングレート・ビット深度統一&頭出し一括同期GUIツール(Windows専用・完全オフライン)
Detect and repair score/audio misalignment in UltraStar karaoke songs
Intelligent media stitch & split toolkit with multimodal adaptive alignment, smooth overlap blending, and gap compensation.
To associate your repository with the audio-alignment topic, visit your repo's landing page and select "manage topics."