louiskirsch/speechT
An opensource speech-to-text software written in tensorflow
Implements the Wav2Letter architecture with CTC loss for end-to-end acoustic modeling, achieving 8% letter error rate on LibriSpeech. Supports optional KenLM language model integration for improved decoding beyond greedy decoding. Provides CLI tools for preprocessing LibriSpeech data, training with TensorBoard monitoring, live microphone transcription, and model evaluation with pre-trained weights available.
160 stars. No commits in the last 6 months.
Stars
160
Forks
34
Language
Python
License
Apache-2.0
Category
Last pushed
Oct 15, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/louiskirsch/speechT"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
liangstein/Chinese-speech-to-text
Chinese Speech To Text Using Wavenet
Open-Speech-EkStep/vakyansh-models
Open source speech to text models for Indic Languages
Open-Speech-EkStep/vakyansh-wav2vec2-experimentation
Repository containing experimentation platform on how to train, infer on wav2vec2 models.
oliverguhr/wav2vec2-live
A live speech recognition using Facebooks wav2vec 2.0 model.
silversparro/wav2letter.pytorch
A fully convolution-network for speech-to-text, built on pytorch.