Kaustuv Kanti Ganguli

Postdoctoral Associate in Machine Learning Affiliation: NYU Abu Dhabi
Education: PhD Indian Institute of Technology Bombay

Research Areas: Computational Musicology, Artificial Intelligence in Music, Music Perception and Cognition, Music Information Retrieval


Dr. Kaustuv Kanti Ganguli is a Postdoctoral Associate in Machine Learning and a 'Scholar in Residence at the Center for Digital Scholarship (CDS) at the New York University Abu Dhabi (NYUAD) in the UAE. His area of expertise includes Artificial Intelligence, Machine Learning, Virtual Reality, Sound Engineering, Audio and Speech Processing, Natural Language Processing, Music Performance and Musicology, Cognitive Neuroscience, Computational Linguistics. Kaustuv holds an administrative position at NYUAD as the Chair of the Postdoctoral Council Steering Committee (PCSC) and is a member of the Academic Strategy Task Force (ASTF), Research Kitchen on Heritage, Memory, and Mobility (HMM). Kaustuv has his startup ‘MuSeeing’ incubated at the StartAD entrepreneurship program.

Kaustuv obtained his PhD degree in Electrical Engineering from the Indian Institute of Technology (IIT) Bombay, India with a best-thesis award entitled "A corpus-based approach to the computational modeling of melody in raga music". He has more than 38 publications including high impact factor journals like JASA, JNMR and got more than 230 citations with h-index: 9, i10-index: 9. Prior to his PhD, Kaustuv secured the University Gold Medal for his undergraduate Bachelor of Technology degree in Instrumentation Engineering with a CGPA of 9.3/10.

Kaustuv has taught several courses in Engineering and Computer Science departments at IIT as well as co-supervised many undergraduate and master's students and capstone projects. His courses stress fundamental concepts and encourage students to be able to ask novel questions. Some of the courses include Advance Topics in Signal Processing, Speech and Audio Coding, Automatic Speech Recognition, Music Technology, Audio Signal Processing for Music Applications, Advanced Musical Programming. Kaustuv delivers lectures on bridge courses between Music Program, Computer Science, and Engineering in the Sound and Music Computing major.

Kaustuv’s research has focused on engineering approaches to build computational models and artificial intelligence on non-Eurogenetic music repertoires. The models are not only suitable for better understanding, preservation, cataloging, and educational goals with the music but are also coherent with human perception and cognition. Kaustuv leads on the front of the computational science in the Music and Sound Cultures (MaSC) research lab where the aim is to study music from the region of the Arabian Gulf through an interdisciplinary understanding of the art and science of music. Some of the recent efforts have shown potential in automatic recognition of Maqam from audio melodies, finding ‘similar’ performances, and visualize an archive through an interactive interface in virtual reality (VR).

Kaustuv is also highly trained in vocal music in the north Indian (Hindustani) tradition for over 28 years. He is a recipient of several prestigious awards and Government accolades including the President’s Gold Medal. He is an active performer and has traveled the globe with his music performance along with scientific conferences. Kaustuv is also a visiting guru at the Malhaar school of performing arts in Dubai and has performed in several concerts sponsored by Emirates NBD and Petrochem Middle East. 


Research

I lead the computational sciences and machine learning tasks at the Music and Sound Cultures (MaSC) research laboratory. We address musicologically interesting problems with techniques from artificial intelligence and music cognition to add knowledge constraints to the data-driven models. Our corpora includes both music from the Gulf and South India.