The field concerned with the analysis, recognition, synthesis and transformation of human speech signals by computational systems.
Semantic Classification
Content
- Speech processing spans tasks including automatic speech recognition, speech synthesis, speaker identification and speech enhancement. It combines signal processing of the acoustic waveform with statistical and neural models that map between audio and linguistic representations.
- Modern systems largely use deep neural networks trained on large speech corpora, often end to end. The field connects acoustics, phonetics and natural language processing, and underpins applications such as voice assistants, transcription and accessibility tools.