zw76859420/ASR_Theory

Curated collection of ASR acoustic modeling approaches built with Kaldi, covering GMM-HMM and NN-HMM architectures across different modeling units (phones, syllables, characters). Includes research papers, presentation slides, and Google's INTERSPEECH 2018 materials on speech recognition theory, with companion implementations demonstrating deep learning architectures for Chinese speech recognition tasks.

619 stars. No commits in the last 6 months.