# microsoft/NeuralSpeech

Repository: https://github.com/microsoft/NeuralSpeech
Canonical: https://ross.abutalabs.com/products/neuralspeech
Language: Python
License: MIT
License Family: permissive
Last push: 2024-02-11T13:20:00+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1763, "days_push": 934, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1462, forks 183 (observed 2026-08-28T04:04:47.605884+00:00)

## What it is
NeuralSpeech is a Microsoft Research Asia research repository containing implementations of neural speech processing models across ASR error correction, text-to-speech synthesis, spatial audio synthesis, and video dubbing. It accompanies peer-reviewed papers published at venues like NeurIPS, ICLR, EMNLP, and ICASSP.

## Use cases
- correct errors in ASR transcriptions with neural models
- synthesize speech from text with lightweight TTS models
- generate binaural spatial audio from mono audio
- dub videos with voice conversion
- reproduce research papers on speech processing
- experiment with prompt-based controllable TTS

## When to choose
- you need reference implementations of published speech research models
- you are a researcher exploring ASR error correction or TTS
- you want to build on Microsoft's speech synthesis and recognition experiments

## When to avoid
- you need a production-ready ASR or TTS service
- you want a maintained end-user tool with documentation and support
- you need a single unified toolkit rather than separate research codebases

## Facets
- artifact type: library
- maturity: active
- function: speech-recognition, tts, audio-processing, machine-learning, deep-learning
- domain: speech-processing, artificial-intelligence, machine-learning
- platform: python, cross-platform
- tags: asr, text-to-speech, research, microsoft, spatial-audio, video-dubbing, error-correction, audio

## Member repositories
- microsoft/NeuralSpeech (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:47.605884+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:35:18.201433+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/NeuralSpeech (fetched 2026-08-28T04:04:47.605884+00:00, sha 28c98f70738b)
- Data as of 2026-08-30T08:39:29.467469+00:00.