Abstract

Technological innovations in the field of speech technologies have rapidly increased in the last few years. However, new technologies cannot be applied directly since they need to be developed and they usually co-exist with the old before supplanting them. In SRI, we transit these innovations from basic research to product development until they are incorporated into society. This technological transition considers user practices, regulations, industrial networks, and infrastructure. We present different protocols and solutions to evaluate, develop and deploy the speech technologies into production.

Biography

Diego Castán, Ph.D., is a senior computer scientist in the Speech Technology and Research Laboratory in SRI International. He has research interests in speaker identification, language identification, multimedia event detection, audio content analysis, audio segmentation and speech activity detection.
At SRI, Castán has contributed to SRI’s Language Identification and Speaker Identification systems for many Defense Advanced Research Projects Agency (DARPA) programs and he has been actively involved in the development of the Open Language Interface for Voice Exploitation (OLIVE).
Prior to joining SRI, Castán was a postdoctoral researcher at the University of Zaragoza, where he focused on speech activity detection and acoustic concepts recognition.
Castán has published more than 40 papers in the field of speech technologies.