Sagar Dutta

Post Doctoral Fellow, RITMO Centre for Interdisciplinary Studies in Rhythm, Time and Motion, University of Oslo, Norway

prof_pic.jpg

Department of Musicology,

University of Oslo

Oslo, Norway

I received my Ph.D. in Electronics and Communication Engineering from the National Institute of Technology Silchar, India (2022). My dissertation, Pattern Recognition Based on Microwave Signal Using Artificial Intelligence, focused on pattern recognition using antenna reactive fields, with applications in human activity recognition (HAR) and machine health monitoring.

Following my Ph.D., I joined the Madhav Lab at the Indian Institute of Technology Kanpur as a Post-Doctoral Fellow, where I worked at the intersection of machine learning and audio signal processing. Drawing on my training as a Western classical guitarist, my research naturally shifted towards music information retrieval (MIR). At IIT Kanpur, I developed and deployed an audio-based recommendation system for India’s national broadcaster, Prasar Bharati, leading the integration of AI models for large-scale multimedia search and recommendation.

I am currently a Post-Doctoral Fellow at the RITMO Centre for Interdisciplinary Studies in Rhythm, Time and Motion, University of Oslo, Norway. My present research focuses on rhythmic analysis of music and dance, with an emphasis on multimodal synchronization. Specifically, I am working on learning-based methods for estimating dance tempo directly from motion-capture data, developing adaptive algorithms that integrate rhythmic information across multiple body segments to achieve state-of-the-art accuracy across diverse dance genres.

Research Interest:

My research lies at the intersection of machine learning, audio signal processing, and music information retrieval, with a special focus on rhythmic analysis of music and dance. I develop learning-based methodologies for audio and movement analysis, exploring multimodal synchronization between music and the body. Currently, I am investigating approaches that integrate motion-capture data with musical structure, advancing applications in dance information retrieval and multimodal interaction.

news

Sep 1, 2024 Paper titled “AudioNet: Supervised Deep Hashing for Retrieval and Ranking of Audio Events” accepted in IEEE Transactions on Audio, Speech and Language Processing

selected publications

  1. Classification of lower limb activities based on discrete wavelet transform using on-body creeping wave propagation
    Sagar Dutta, Banani Basu, and Fazal Ahmed Talukdar
    IEEE Transactions on Instrumentation and Measurement, 2020
  2. Classification of induction motor fault and imbalance based on vibration signal using single antenna’s reactive near field
    Sagar Dutta, Banani Basu, and Fazal Ahmed Talukdar
    IEEE Transactions on Instrumentation and Measurement, 2021
  3. Classification of motor faults based on transmission coefficient and reflection coefficient of omni-directional antenna using DCNN
    Sagar Dutta, Banani Basu, and Fazal Ahmed Talukdar
    Expert Systems with Applications, 2022
  4. Cascaded neural network based small array synthesis with robustness to noise
    Sagar Dutta, Banani Basu, and Fazal Ahmed Talukdar
    International Journal of RF and Microwave Computer-Aided Engineering, 2021
  5. Classification of scattering parameters of body-embedded wideband textile antennas for early diagnosis and monitoring of breast cancer
    Nirmalya Das, Banani Basu, Sagar Dutta, and 1 more author
    International Journal of Microwave and Wireless Technologies, 2023