Dipesh Tamboli

I am currently a third PhD student at Purdue University, working with Prof. Vaneet Aggarwal in the field of Generative AI and its safety.

Last year, I worked at Amazon Robotics as an Applied Scientist. I hope to attend Meta (Foundations Model Team) for my Summer 2024 Internship.

Previously, I was working as a Machine Learning Researcher at AWL Inc., researching on self-supervised learning for object detection.

I finished my undergrad from IIT Bombay with a major in Electrical Engineering and a Minor in Computer Science. I was advised by Prof. Biplab Banerjee and Prof. Subhasis Chaudhuri for my Bachelor Thesis Project.

During my undergrad, I have interned at AWL Inc., Video Analytics Laboratory (VAL) at IISc Banglore under the supervision of Prof. R. Venkatesh Babu, with Prof. Pengtao Xie at University of California San Diego, and with Prof. Fabio Cuzzolin at VAIL Labs, Oxford.

Email  /  Resume  /  Google Scholar  /  LinkedIn  /  Github

I'm interested in computer vision, machine learning, optimization, and image processing. Please check my Google Scholar for the recent publications.

We propose a novel learning setting for multi-source domain adaptation where the target-domain may contain open-set classes.

We propose a data-free method of extracting Impressions of each class from the classifier's memory.

We tackle the problem of image inpainting in the remote sensing domain. We propose a novel inpainting method that individually focuses on each aspect of an image such as edges, colour and texture using a task specific GAN.

We propose a deep learning (DL) architecture dubbed bidirectional autoencoder for nanophotonic metasurface design via a template search methodology.

We aim to provide a better interpretation of classification results by providing localization on microscopic histopathology images. We frame the image classification problem as weakly supervised multiple instance learning problem where an image is collection of patches i.e., instances.

We trained YOLO for leaves tips detection and to estimate the number of leaves in a rice plant. With this proposed framework, we screened the genotypes based on selected traits. These genotypes were further grouped among different groupings of drought-tolerant and drought susceptible genotypes using the Ward method of clustering.