.jpg)
D I P A M P A U L
>> Research Assistant @ Carnegie Mellon University, Emory University
>> Former Research Intern @ Georgia Tech, Indian Institute of Technology, Bombay
About Me
>> Primary Research Interests: Deep Learning, Computer Vision, Medical Imaging, Representation Learning, and Intelligence Theory.
>> Secondary Research Interests: Computational Game theory, and Heuristic Algorithms.
My name is Dipam Paul and I am a final year undergraduate from KIIT University, India pursuing Electronics and Telecommunication Engineering and bearing a current GPA of 9.15/10. I am currently working at Xu Lab @ Carnegie Mellon University working on areas of Object Detection of macromolecules concerning 3D Cryo-ET Data and Deep Active Learning with a special focus on solving the Catastrophic Interference problem in Dynamically Expandable Networks. We are wrapping this up and this present week is my last working week here.
I have also worked at Banerjee Lab @ Emory University as a Research Assistant and we worked on a problem to solve the problems associated with Deep Brain Stimulation (DBS) for patients with Parkinson's disease and got our work on Super Learner published (short paper) at Trustworthy AI for Healthcare workshop @ AAAI'21 under the supervision of Dr. Imon Banerjee. We have also recently submitted two workshop papers (under review) at CVPR 2021 and two accepted papers at four workshops at ICLR 2021. Recently, I have also served as a reviewer at the Science and Engineering of Deep Learning workshop at ICLR 2021.
Before this, I worked as a Research Intern till 14th August 2020 at Dr. Eva Dyer's lab (NerDS) at Georgia Tech in the area of Computational Neuroscience. I worked on a Triplet Loss ranking problem to embed Brain Imagery while also creating a ranking dataset for the same.
Previously, I have also worked at Dr. Amit Sethi's MeDAL Lab at IIT Bombay which encompasses and works on areas of Biomedical Imaging and Analysis. We are currently working on modeling and training an Auxiliary Classifier GAN to optimize the latent vector variables and hence improve on a model to detect COVID-19 from Chest X-Ray Images (Corona-NET), while also working to integrate a semi-supervised approach into the same using SimCLR to counteract the lack of availability of data.
Currently, I am particularly interested in developing AI tools in the area of Continual Learning, and Deep Generative Models, Trustworthy AI through the means of working with esoteric ranges of synthetic and non-synthetic Data and the manifestation of the same into understanding 'Intelligence' at a theoretical and intuitive level.