Diego Castán

Advanced Computer Scientist, Speech Technology and Research Laboratory (STAR)

Diego Castán, Ph.D., has research interests in multimedia event detection, audio content analysis, audio segmentation and speech activity detection.

At SRI, Castán is contributing to SRI’s Language Identification and Speaker Identification systems for the Defense Advanced Research Projects Agency (DARPA) Robust Automatic Transcription of Speech (RATS) program. He was actively involved in the multimedia event detection task for the Intelligence Advanced Research Projects Activity (IARPA) Automated Low-Level Analysis and Description of Diverse Intelligence Video (ALADDIN) program as international fellow during 2012.

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 15 papers in the field of speech technologies. View selected publications on Google Scholar.

Recent publications

more +

Toward Fail-Safe Speaker Recognition: Trial-Based Calibration with a Reject Option (1/1/2019) - The output scores of most speaker recognition systems are not directly interpretable as stand-alone values. For this reason, a calibration…

Analysis of Complementary Information Sources in the Speaker Embeddings Framework (9/1/2018) - Deep neural network (DNN)-based speaker embeddings have resulted in new, state-of-the-art text-independent speaker recognition technology. However, very limited effort has…

How to train your speaker embedding extractor (6/1/2018) - With the recent introduction of speaker embeddings for text-independent speaker recognition, many fundamental questions require addressing in order to fast-track…