Sound to Sense: Introduction to the Special Session

Sound to Sense (S2S) is a Marie Curie Research Training Network, funded 2007-2011. It involves some 50 researchers in 13 institutions in 10 countries. The ultimate aim of S2S is to provide models of speech processing that closely reflect the exquisite flexibility and robustness of human speech processing (HSP), that pave the way for the next generation of robust automatic speech recognition (ASR) and text-to-speech (TTS) machines, and that promise a new theoretical basis for foreign language (FL) teaching, and diagnosis and treatment of speech disorders. The immediate aim is to elucidate the interaction of knowledge and sensation in speech perception, using insights from recent linguistic, phonetic and psychological research to inform speech recognition models (HSP and ASR). The new models should better reflect the way humans listen and respond to their native language (L1) and to FLs. One focus is to track how phonetic information, especially fine phonetic detail that varies systematically with linguistic and interactional structure and function, is used in differing situations: when listeners have appropriate linguistic-phonetic knowledge (listening to L1), inadequate or inappropriate knowledge (listening to FL), and inadequate access to the signal (listening in adverse conditions). A related focus is to elucidate how phonetic information contributes to understanding. S2S aims both to combine knowledge from independent disciplines and to reduce fragmentation within disciplines. Young research workers, with multidisciplinary training focused on a new theoretical framework, will acquire the perspective and skills needed to allow them to take speech processing research and applications further than any one discipline can currently achieve.