ASRT_SpeechRecognition and ctc-asr
Tool A, a deep-learning-based Chinese speech recognition system, and Tool B, an end-to-end trained speech recognition system based on RNNs and CTC, are ecosystem siblings because Tool A is an implementation of a CTC-based ASR system, demonstrating the practical application of the concepts and techniques that Tool B describes in its abstract.
About ASRT_SpeechRecognition
nl8590687/ASRT_SpeechRecognition
A Deep-Learning-Based Chinese Speech Recognition System 基于深度学习的中文语音识别系统
Combines deep convolutional neural networks with LSTM and attention mechanisms using CTC loss for acoustic modeling, paired with a maximum-entropy hidden Markov model for converting pinyin sequences to Chinese characters. Supports multi-dataset training (THCHS30, AIShell-1, MagicData, etc.) and deploys via HTTP/gRPC APIs with client SDKs for Windows, Python, Go, and Java; achieves ~85% pinyin accuracy on test sets with audio inputs up to 16 seconds.
About ctc-asr
mdangschat/ctc-asr
End-to-end trained speech recognition system, based on RNNs and the connectionist temporal classification (CTC) cost function.
Implements bidirectional RNN layers with dense layers trained on 900+ hours of multi-corpus audio data (LibriSpeech, Common Voice, TEDLIUM, Tatoeba), achieving 12.6% WER without external language models. Built on TensorFlow with configurable architecture parameters, supporting GPU acceleration and modular training/evaluation workflows via CSV-based corpus definitions. Includes utilities for multi-corpus preparation, checkpoint management, and real-time training visualization through TensorBoard.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work