Tanvina Patel is a Post Doctoral Researcher at the Multimedia Computing Group at Delft University of Technology, the Netherlands. She works on Inclusive ASR and aims to improve usability of ASR systems. Her other research interests include speech signal analysis, and anti-spoofing for voice-biometrics.
Prior to her Post-doc she was a Data-Scientist at Cogknit Semantics Pvt. Ltd., Bangalore for more than 4years. At Cogknit she was associated with developing speech recognition systems (for both data rich and low-resourced languages) and various speech-technology applications that enhance human-computer interaction and solve market problems.
She earned her Ph.D. in April 2017 from DA-IICT, Gandhinagar. During 2012-17, she was associated with two DeitY, GoI sponsored consortium projects on TTS and ASR in Indian Languages. She has received the Int.l Speech Comm. Assoc. (ISCA), Microsoft Research India (MSR), and IEEE Signal Processing Society (SPS) Travel Grant to present research papers at INTERSPEECH'13, INTERSPEECH'14, and ICASSP'16, respectively.
Her selective achievements includes Cogknit's ASR submission that was ranked second at the Microsoft Low Resource Challenge in Indian Languages, INTERSPEECH'18. As a part of her Ph.D she participated and developed the best performing countermeasure to detect Natural vs. Spoofed speech in 'ASV Spoof 2015 Challenge' held at INTERSPEECH 2015, Dresden, Germany. She holds 20+ publications and also is/was a reviewer for INTERSPEECH, ICASSP, Computer Speech & Language (Elsevier), Speech Communication and Transactions on Information Forensics & Security.
For other info and detailed publications: Personal Webpage: https://sites.google.com/site/tanvinabpatel/