معرفی
Dr. Arshdeep Singh is a Research Fellow A at the Centre for Vision, Speech and Signal Processing (CVSSP) at the University of Surrey, specializing in AI-driven audio processing. He concurrently serves as a Sustainability Fellow at the University’s Institute for Sustainability and holds the UKAN+ Early Career Acoustic Champion title (2023). His research focuses on efficient neural network architectures for audio classification, model compression, and sustainable AI applications. Singh completed his PhD at IIT Mandi, India, with prior roles including internships at Intel Bangalore and CSIR-CSIO Chandigarh, and an M.E. from Panjab University (Gold Medal recipient).
Research interests include audio signal processing, CNN compression via filter pruning, and sustainable AI. Notable contributions include developing lightweight CNN frameworks for acoustic scene classification and sound event detection, with applications in smart home technologies and industrial health monitoring. His work often targets resource-constrained devices, emphasizing real-world deployability.
Publications highlight innovations like graph centrality-based filter pruning and quaternion transformations for CNN efficiency. Awards include the Panjab University Gold Medal for M.E. and recognition as an Early Career Acoustic Champion. Supervision includes MSc students (e.g., Soham Bhattacharya, Bars Szegedi) and undergraduates (Kristaps Redmers, Yan-Ping Liao). His interdisciplinary projects span audio captioning, embedded hardware deployment, and dataset creation for sound event detection in residential environments.
