Hello, I'm
Dr. Satwinder Singh
Specializing in inclusive AI, speech recognition for atypical speech, and accessible language technologies. Passionate about bridging the gap between advanced machine learning and real-world clinical applications.
LinkedIn
Get To Know More
About Me
Experience
7+ years
Research & Teaching
Education
PhD Computer Science
Massey University, NZ
I work on inclusive AI for speech and language technologies, with a focus on automatic speech recognition for atypical speech, low-resource settings, and accessibility-driven systems. I also supervise student projects and collaborate with clinicians and community partners to ensure research impact.
Explore My
Experience & Skills
Technical Skills
Python
PyTorch
TensorFlow
C/C++
Linux/Bash
SQL
Git
Docker
Web Dev
Research Expertise
Speech Recognition (ASR)
Deep Learning
NLP
AI Accessibility
Low-resource Languages
Atypical Speech
Human-Computer Interaction
Relevant Experience
Postdoctoral Research Fellow
Dept. of Electrical, Computer and Software Engineering, University of Auckland | Auckland, NZ
Jan 2024 - Jan 2026
- Working on Automatic Speech Recognition for Dysarthric Speech.
- Supervising undergraduate and postgraduate research.
Postdoctoral Research Fellow
School of Mathematical and Computational Sciences, Massey University | Auckland, NZ
Jul 2023 - Dec 2023
- Worked on Natural language processing for Q&A in indigenous/vernacular languages.
- Project funded by MBIE Catalyst: Strategic – New Zealand-Singapore Data Science Research Programme.
Research Assistant
School of Mathematical and Computational Sciences, Massey University | Auckland, NZ
Nov 2022 - Mar 2023
- Assisted with Course: Information Sciences Research Methods (158750).
- Worked on funding applications such as Marsden and MURF.
- Conducted research and data collection for Catalyst Funding Project.
Tutor
Pinnacle Global Academy (PGA) | Auckland, NZ
Jul 2021 - Feb 2022
- Taught Mathematics and Python programming.
- Prepared students for competitive exams.
Assistant Professor
Faculty of Computational Sciences, GNA University | Punjab, INDIA
Jul 2016 - Feb 2018
- Taught graduate and post-graduate courses.
- Managed lecture scheduling and supervised examinations.
- Supervised students' postgraduate research.
What I've Been Up To
Recent Activities
October 2025
Best Research Awards at Exihibition Day 2025
Part 4 Research students delivered outstanding work across a range of impactful topics and Secured best industry project in image and voice processing catrory 1 and 2.
September 2025
Transformer Models Workshop
Organized and led a one-day workshop on Transformer models with hands-on practical sessions at the University of Auckland.
April 2025
ICASSP 2025 Conference
Presented two papers on dysarthric speech recognition at the IEEE International Conference on Acoustics, Speech and Signal Processing in India.
December 2024
ICONIP 2024 Conference
Presented research on comprehensive performance evaluation of Whisper models in dysarthric speech recognition at ICONIP in New Zealand.
December 2024
Local Arrangement Chair
Served as Local Arrangement Chair for the ACM Multimedia Asia Conference held in Auckland, New Zealand.
November 2024
Best Project Award 2024
Supervised students Steven Li and Adi Shenoy, who won the Best Project Award for developing a Neural Speaker Diarization System for Doctors.
Ongoing
PhD Student Supervision
Currently supervising multiple PhD and Honours students on projects related to dysarthric speech recognition, low-resource ASR, and AI accessibility tools.
Browse My
Recent Publications
2025
Under review
Low-Burden Data Augmentation for Dysarthric ASR via Zero-Shot Voice Cloning
ICASSP
2025
Published
Convolution-Augmented Transformers for Enhanced Speaker-Independent Dysarthric Speech Recognition
IEEE Transactions on Neural Systems and Rehabilitation Engineering
2025
Published
Robust Cross-Etiology and Speaker-Independent Dysarthric Speech Recognition
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), India, 2025
2025
Published
Dysarthric Speech Conformer: Adaptation for Sequence-to-Sequence Dysarthric Speech Recognition
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), India, 2025
2024
Published
Mix-fine-tune: An Alternate Fine-tuning Strategy for Domain Adaptation and Generalization of Low-resource ASR
Proceedings of the 6th ACM International Conference on Multimedia in Asia
2024
Published
A Comprehensive Performance Evaluation of Whisper Models in Dysarthric Speech Recognition
International Conference on Neural Information Processing (ICONIP), New Zealand, 2024
2023
Published
PhasePerturbation: Speech Data Augmentation via Phase Perturbation for Automatic Speech Recognition
ACM MM Asia, 2023
2023
Published
Real and Synthetic Punjabi Speech Datasets for Speech Recognition
Data in Brief Journal, 2023
2023
Published
A Novel Self-training Approach for Low-resource Speech Recognition
Proc. Interspeech, Ireland, pp. 1588-1592, 2023
Explore My
Featured Projects
Active
Dysarthric Speech Recognition System
An end-to-end automatic speech recognition system designed specifically for dysarthric speech using convolution-augmented transformers. Achieves state-of-the-art performance across multiple etiology groups.
2024
Neural Speaker Diarization for Medical Consultations
Supervised development of an AI system to automatically identify and separate speakers in doctor-patient conversations. Developed by students Steven Li and Adi Shenoy, winning Best Project Award in 2024.
Research
Te Reo Māori Speech Recognition
Developing robust ASR models for te reo Māori using self-training and meta-learning approaches to overcome data scarcity. Collaborating with Māori language experts to ensure cultural appropriateness.
2023
Punjabi Speech Corpus
Created comprehensive real and synthetic speech datasets for Punjabi language containing over 100 hours of annotated audio. Publicly available for research purposes on Mendeley Data.
2023
PhasePerturbation Augmentation
Novel speech data augmentation technique using phase perturbation to improve ASR model robustness and generalization without requiring additional labeled data. Published at ACM MM Asia 2023.
Updates
News and Blog
2025-12-21
Scaling Accessibility: How Zero-Shot Voice Cloning Boosts Dysarthric Speech Recognition