Scholar
Desh Raj
Google Scholar ID: Z-7pPYEAAAAJ
Meta AI
Citations & Impact
Citations
1,411
H-index
16
i10-index
22
Publications
20
Co-authors
86
list available
Publications
Resume (English only)
Academic Achievements
- Named IEEE Rising Star in Signal Processing in 2023.
- Awarded Fred Jelinek Fellowship at Johns Hopkins for 2023–24.
- Recipient of inaugural JHU-Amazon AI2AI Fellowship for 2022–23.
- Multiple papers accepted at top-tier conferences, including:
- - 2025: 3 papers at IEEE ICASSP 2025 on SpeechLLMs and multi-channel speech foundation models; a new paper on reasoning in SpeechLLMs from Ian Shih’s internship.
- - 2023: GSS paper accepted at InterSpeech 2023 (used in CHiME-7 DASR challenge baseline); 2 papers at IEEE ICASSP 2023.
- - 2022: 2 papers at IEEE ICASSP 2022 on multi-talker ASR with neural transducers and fine-tuning large self-supervised models.
- - 2021: 4 papers at INTERSPEECH 2021; JHU-GoVivace team placed 2nd overall (1st in Hindi-English task) in Indic code-switching challenge; Hitachi-JHU team achieved 2nd best DER in Third Dihard Challenge.
- - 2020: 4 papers accepted at IEEE SLT 2021; JHU submission ranked second-best in CHiME-6 Track 2 (diarization + ASR).
Research Experience
- Joined Meta (NYC) as a Research Scientist in January 2024, working on robust on-device ASR in the AI Speech & EMG team led by Mike Seltzer.
- Since July 2024, the AI Speech team has been part of GenAI, tasked with developing speech for LLaMA models.
- Interned at Meta AI (Menlo Park) Speech team in summer 2022.
- Interned virtually at Microsoft with Dr. Jinyu Li in summer 2021.
- Participated in JSALT 2020 and JSALT 2023 summer workshops, working on informed target-speaker ASR and WFST+end-to-end speech methods, respectively.
- Participated in the Mini SCALE workshop (January 2022) in the 'Improving speech analytics for room audio' team led by Matthew Wiesner.
