Chris10M/Lip2Speech
A pipeline to read lips and generate speech for the read content, i.e Lip to Speech Synthesis.
This tool helps convert silent video footage of a person speaking into audible speech. It takes a video of someone's face as they form words, analyzes their lip movements, and then generates an audio file of what they are saying. It's designed for researchers or practitioners working with video analysis, accessibility, or voice generation.
No commits in the last 6 months.
Use this if you need to extract spoken words from video where audio is unavailable or unclear, by analyzing the speaker's lip movements.
Not ideal if you already have clear audio or if the speaker's lips are not clearly visible in the video.
Stars
93
Forks
22
Language
Python
License
MIT
Category
Last pushed
Jul 23, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/Chris10M/Lip2Speech"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
primepake/wav2lip_288x288
Wav2Lip version 288 and pipeline to train
SARIT42/lipsyncr
LipSyncr is a lip reading web app based on the LipNet model that can lip read videos.
Markfryazino/wav2lip-hq
Extension of Wav2Lip repository for processing high-quality videos.
d-kavinraja/MouthMap
MouthMap is a deep learning-based lip reading system that converts silent video sequences into...
adhadse/Deepdubpy
A complete end-to-end Deep Learning system to generate high quality human like speech in English...