IITKGP

Research Areas

Our research team is now focussed on strengthening our lead in two application domains related to audio processing: (i)Use of speech for biometric authentication and development of counter measure to prevent voice based identity hacking; (ii)Development of heart and lung sound based diagnostic tools for physician's aid. In neuro-signal processing, the team extends our work on how brain works for complex cognitive tasks like perception, imagination, preference decision etc. Going forward, we would like to explore how integration of these three verticals open up newer understanding and newers solutions. We also take up sub-problems leveraging our strengths in these to attend specific R & D needs. In each of the three areas of investigation, we have collaborating scientists, medical experts extending necessary support to our team.
  • Neural style transfer architectures for improving generalization in low-resource spoken language identification by Dey S., Saha G. Engineering Applications of Artificial Intelligence 167 - (2026)
  • System and method for automatic synthetic speech detection for speech-based biometric authentication. by Saha G., Paul D. , Pal M. - (2024)
  • Analysis and classification of beat-level ECG arrhythmia using WST-inspired CNN framework by Nahak S., Saha G. Biomedical Signal Processing and Control 109 - (2025)
  • Estimation of lung sound cycle span using spectro-temporal respiratory frequency evaluation by Bandyopadhyaya I., Singh P., Nahak S., Maity A., Saha G. Applied Acoustics 229 - (2025)
  • Probing Layer-Wise Self-Supervised Representations for Low-Resource Spoken Language Identification by Dey S., Saha G. IEEE Journal on Selected Topics in Signal Processing 19 1436-1447 (2025)
  • A Deep Learning-Based Study on Inter-Language Confusion Assessment Using Spoken Language Identification for Indian Languages by Rajak B., Sinha S., Dey S., Saha G. Proceedings of the National Conference on Communications, NCC - (2025)
  • Non-Invasive Detection of Coronary Artery Disease and Valvular Disorders Using a Multichannel PCG Vest by Fynn M., Marocchi M., Maity A., Mandana K., Rashid J., Rong Y., Saha G. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS - (2025)
  • Enhancing cross-domain robustness in phonocardiogram signal classification using domain-invariant preprocessing and transfer learning by Maity A., Saha G. Computer Methods and Programs in Biomedicine 257 - (2024)
  • Portable Doppler Ultra Sound System for Automatic Identification of Blood Flow Related Diseases by Jana B., Biswas R. , Banerjee S. , Saha G. - (2025)

Principal Investigator

  • A Portable Multisensor Device for Work of Breathing Monitoring, Acquisition and Analysis to Detect Severity of Chronic Respiratory Diseases IIT KHARAGPUR AI4ICPS I HUB FOUNDATION
  • Cost-effective and high precision weight measurement system for non ambulatory sick adult and pediatric patients INDIAN COUNCIL OF MEDICAL RESEARCH (ICMR)
  • Real Time Voice Deepfake Detection System IndiaAl
  • Smart Medical Care System for Safer Transportation of Patients in Rural Area INDIAN COUNCIL OF MEDICAL RESEARCH (ICMR)
  • Use of AI and multichannel, non-invasive sensing for affordable diagnosis of heart diseases. Scheme for Promotion of Academic and Research Collaboration (SPARC), Apex Committee of SPARC, Ministry of Education

Ph. D. Students

Anushka Ghosh

Area of Research: Speech / Audio Processing

Shalini Mukhopadhyay

Area of Research: Biomedical Signal Processing

Souvik Sinha

Area of Research: Audio Signal Processing

Vivek Pratap Singh

Area of Research: Biomedical