C++ based example console application that consumes KaldiService REST API functions for processing audio files and performing speech recognition.

FastCGI support for Kaldi ASR

CMUSphinx based ASR with Qt C++ User interface

MLPerf RNN-T Inference Development

This is now the official location of the Kaldi project.

FastCGI support for Kaldi ASR

Daktilograf STT - Je softver za pretvaranje glasa u tekst koristenjem najsavremenije tehnologije vještačke inteligencije. Fokus nam je na slovenkim jezicima za sad Srpski, Hrvatski, Bosanksi i Crnogorski, u beta verziji je Ruski a pripremamo Poljski i Ukrainski.

High-performance GPGPU inference of OpenAI's Whisper automatic speech recognition (ASR) model

Kaldi-ASR powered by PyTorch C++ API (Experimental)

A project based on https://github.com/alphacep/vosk-api, add necessary kaldi source files to make compiling easier.

Media Agent

Matrix Voice streaming via WebSockets.

shell,python,cpp,CS310,nlp,asr

Calculates the word error rate between the reference and hypothesis in ASR, then print the aligned result.

C++ Client Library for Gliadia API

ICASSP 2020

Rust wrapper for microsoft speech recognition

This project reveals full Respeaker Core V2 potential by using bundled Alango DSP algorithms for audio pre-processing before sending to the custom ASR server.

ASR online decoding using Kaldi NNet3 GrammarFST