Mainak Singha
Mainak Singha
Mainak Singha is a Ph.D. candidate in the Multimedia and Human Understanding Group at the Department of Information Engineering and Computer Science, University of Trento, Italy and a researcher within the ANT MSCA Doctoral Network, supervised by Prof. Elisa Ricci & Prof. Paolo Casari (University of Trento), Prof. Jie Yang (TU Delft) and Dr. Frans Widdershoven (NXP Semiconductors). His research primarily focuses on federated learning, mixture-of-experts and model calibration for efficient, reliable and trustworthy multimodal large vision-language models (LVLMs).
Before joining Ph.D., he was an AI Research Scientist at Tokyo Research Center (TRC) of Aisin Corporation, in Tokyo, Japan, and worked on the 3D vision perception and resource-constrained AI devices for Autonomous Driving applications. He also received a Master’s degree from the Indian Institute of Technology Bombay, India, where he worked on the closed and open-world domain generalization and adaptation of vision-language foundation models. During his Master’s studies, he also did two research internships at MavenAI Technologies as a Machine Learning Intern and at Qen Labs Inc. as AI Research/Consultant Intern.
From the beginning of his research journey, he regularly publishes in top-tier computer vision venues including CVPR, ICCV, ECCV etc, and serves as a committee member in multiple top-tier conferences and journals. He is also the recipient of two Best Paper Awards in ICVGIP’23 and ECML-PKDDw’23. For further details, please visit – https://mainaksingha01.github.io .