Reaching over the gap: A review of efforts to link human and automatic speech recognition research

Abstract

The fields of human speech recognition (HSR) and automatic speech recognition (ASR) both investigate parts of the speech recognition process and have word recognition as their central issue. Although the research fields appear closely related, their aims and research methods are quite different. Despite these differences there is, however, lately a growing interest in possible cross-fertilisation. Researchers from both ASR and HSR are realising the potential benefit of looking at the research field on the other side of the ‘gap’. In this paper, we provide an overview of past and present efforts to link human and automatic speech recognition research and present an overview of the literature describing the performance difference between machines and human listeners. The focus of the paper is on the mutual benefits to be derived from establishing closer collaborations and knowledge interchange between ASR and HSR. The paper ends with an argument for more and closer collaborations between researchers of ASR and HSR to further improve research in both fields

An algorithm that improves speech intelligibility in noise for normal-hearing listeners

“…Despite these advances, human listener's word error rates are often more than an order of magnitude lower than those of state-of-the art recognizers in both quiet and degraded environments ͑Lippmann, 1997, Sroka and Braida, 2005;Scharenborg, 2007͒. Large advances have also been made on the development of algorithms that suppress noise without introducing much distortion to the speech signal ͑Loizou, 2007͒.…”

Towards a neurocomputational model of speech production and perception

“…Benzeghiba et al 2007, Scharenborg 2007. Current corpus-based speech synthesis systems are limited as well, especially concerning (i) flexibility in modeling different speaker and voice characteristics and concerning (ii) segmental as well as prosodic naturalness (e.g.…”

Modeling the use of durational information in human spoken-word recognition

“…To that end, Fine-Tracker is built using techniques from the field of automatic speech recogni tion (ASR), and as such is part of a growing line of research aimed at bridging the research fields of psycholinguistics and ASR (for an overview, see Scharenborg, 2007).…”