an open-source implementation of sequence-to-sequence based speech processing engine

Rust re-implementation of OpenFST - library for constructing, combining, optimizing, and searching weighted finite-state transducers (FSTs). A Python binding is also available.

Port of the OpenFST library to Windows

ASR for Chinese Mandarin

Source code of the TinyOWS project. Please submit pull requests to the 'main' branch.

🔥 ASR教程: https://dataxujing.github.io/ASR-paper/

A lightweight javascript module to format WFS-T-2 statements from GeoJSON features

A high-performance inference and serving framework for Automatic Speech Recognition

Play with Weighted Finite State Transducers (WFST) in the Julia language.

A sample application with a React/OpenLayers frontend that writes feature data to PostGIS using GeoServer WFS Transaction support.

A Julia recipe for training an ASR system using the TIDIGITS database

Notebooks for "AI 응용-음성"

Post processing for speech recognition

基于 WFST 与 GNN 的物联网侧信道攻击框架，研究 IFTTT/TAP 智能家居自动化规则的隐私泄露风险 | IoT WFST side-channel attack framework for evaluating smart-home rule privacy leakage on Trigger-Action Platforms

Some very old scripts (about 4 years ago) written by myself to convert ARPA language model, HTK lattice, etc to weighted finite state transducers.

WFST-based ASR system with custom Viterbi decoder — 81% WER reduction through adaptive pruning