Aadya Arora

Junior undergraduate at the Indian Institute of Technology Gandhinagar, majoring in Electrical Engineering with a minor in Artificial Intelligence. I focus on leveraging Machine Learning, Deep Learning, and Computer Vision to solve complex challenges.

Currently preparing to join Microsoft R&D as a Data Science Intern in Summer 2025. Previously, I worked on medical imaging projects under Dr. Vinay Namboodiri (University of Bath) and shadow segmentation under Dr. Shanmuganathan Raman (IIT Gandhinagar).

Email  /  Google Scholar  /  GitHub  /  LinkedIn

With the popularity of foundational models, parameter efficient fine tuning has become the defacto approach to leverage pretrained models to perform downstream tasks. Taking inspiration from recent advances in large language models, Visual Prompt Tuning, and similar techniques, learn an additional prompt to efficiently finetune a pretrained vision foundational model. However, we observe that such prompting is insufficient for fine-grained visual classification tasks such as medical image classification, where there is large inter-class variance, and small intra-class variance. Hence, in this paper we propose to leverage advanced segmentation capabilities of Segment Anything Model 2 (SAM2) as a visual prompting cue to help visual encoder in the CLIP (Contrastive Language-Image Pretraining) by guiding the attention in CLIP visual encoder to relevant regions in the image. This helps the model to focus on highly discriminative regions, without getting distracted from visually similar background features, an essential requirement in a fewshot, finegrained classification setting. We evaluate our method on diverse medical datasets including X-rays, CT scans, and MRI images, and report an accuracy of (71%, 81%, 86%, 58%) from the proposed approach on (COVID, lung-disease, brain-tumor, breast-cancer) datasets against (66\%, 70\%, 68\%, 29\%) from a pretrained CLIP model after fewshot training. The proposed approach also allows to obtain interpretable explanation for the classification performance through the localization obtained using segmentation.

Shadow removal and segmentation remain challenging tasks in computer vision, particularly in complex real-world scenarios. This study presents a novel approach that enhances the ShadowFormer model by incorporating Masked Autoencoder (MAE) priors and Fast Fourier Convolution (FFC) blocks, leading to significantly faster convergence and improved performance. We introduce key innovations: (1) integration of MAE priors trained on Places2 dataset for better context understanding, (2) adoption of Haar wavelet features for enhanced edge detection and multi-scale analysis, and (3) implementation of a modified SAM Adapter for robust shadow segmentation. Extensive experiments on the challenging DESOBA dataset demonstrate that our approach achieves state-of-the-art results, with notable improvements in both convergence speed and shadow removal quality.