Scholar
Nithin Rao Koluguri
Google Scholar ID: YPtcUTUAAAAJ
NVIDIA Corporation
Speech Processing
Deep Neural Networks
Machine Learning
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
768
H-index
14
i10-index
16
Publications
20
Co-authors
14
list available
Contact
Email
nithinrao.koluguri@gmail.com
CV
Open ↗
GitHub
Open ↗
LinkedIn
Open ↗
Publications
4 items
Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST
2025
Cited
0
Speaker Targeting via Self-Speaker Adaptation for Multi-talker ASR
2025
Cited
0
Training and Inference Efficiency of Encoder-Decoder Speech Models
2025
Cited
0
Sortformer: Seamless Integration of Speaker Diarization and ASR by Bridging Timestamps and Tokens
arXiv.org · 2024
Cited
5
Resume (English only)
Academic Achievements
Created TitaNet architecture for speaker recognition, widely adopted with ~1.5M monthly downloads on Hugging Face
Co-built the first speaker diarization modules in NeMo using TitaNet embeddings and NME-SC clustering
Led billion-parameter scaling of FastConformer ASR models to advance speech recognition capabilities
Leading development of Parakeet model series; parakeet-tdt-0.6b-v2 currently ranks #1 on Hugging Face Open-ASR leaderboard
Co-author of 'Sortformer: Seamless integration of speaker diarization and ASR by bridging timestamps and tokens'
Developed speech classifiers for ALS and Parkinson’s disease detection using voice as a biomarker at IISc
Improved ASR by generating n-best path lists under various noise conditions using Kaldi during USC research
Co-authors
14 total
Boris Ginsburg
NVIDIA
Taejin Park
NVIDIA
He Huang
NVIDIA
Kunal Dhawan
Research Scientist, NVIDIA
Krishna C. Puvvada
NVIDIA
Oleksii Hrinchuk
NVIDIA
Somshubra Majumdar
NVIDIA
Vitaly Lavrukhin
NVIDIA
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up