|
|
|
I am an undergraduate student majoring in Economics and Computer Science, with a minor in Data Science, at BITS Pilani, Goa. My interests lie in exploring cutting-edge research and applying computer vision techniques to solve real-world problems across diverse domains. I have explored areas such as drone-based navigation in unknown indoor environments, efficient data selection through coreset methodologies, and the use of Large Language Models (LLMs) for generating insightful academic reviews. My current focus lies in pushing the boundaries of pose tracking and 3D scene generation, where I aim to bridge the gap between textual descriptions and visually rich outputs. During the summer of '23, I had the privilege of serving as a research intern in the Integrated Circuit and Systems Group at CSIR CEERI Pilani, supervised by Dr. Kaushal Kishore. My work centered on training and implementing algorithms, specifically the Continuous Actor Critic Learning Automata for Autonomous Indoor Drone Navigation. The culmination of our efforts involved simulating this algorithm in the Gazebo environment. And in the summer of '24, I worked on Data efficiency methods for GAN Training and other Computer Vision Tasks at IIT Hyderabad. Feel free to drop me an e-mail if you want to chat with me! |
|
July '24 |
Selected for DataLab Project under supervision of Professor Ashwin Srinivasan. |
|
June '24 |
Selected for SURE IIT Hyderabad 2024 Program to work at DiL Lab under Professor Konda Reddy Mopuri. |
|
May '23 |
Started Research Internship at CEERI under the supervision of Dr. Kaushal Kishore. |
|
Jan '23 |
I will be a Teaching Assistant for ECON F412 : Security Analysis and Portfolio Management at BITS Pilani, Goa. |
|
June '24 - Aug '24 Working on Data efficiency methods for Computer Vision Tasks. |
|
May '23 - May '24 Working on Algorithms for Autonomous Indoor Drone Navigation in Integrated Circuit and Systems Group under the supervision of Dr. Kaushal Kishore. |
|
|
Under Dr. Kaushal Kishore
This project explores the implementation of Continuous Actor Critic Learning Automata (CACLA) to achieve autonomous indoor drone navigation within a simulated Gazebo environment using ROS2. The project leverages the CACLA algorithm, as proposed in this research paper, to enable the drone to learn and adapt its navigation strategies in real-time within unknown indoor spaces. |
|
This project investigates the use of transformers for video classification in the UCF101 dataset, focusing on the top 5 categories for enhanced performance and efficiency. It employs statistical tests, explores transformer architectures, optimizes parameters using techniques like mathematical morphology, and comprehensively evaluates the model's accuracy and generalizability. |
|
Embraced the magic of Neural Radiance Fields (NeRF)! I delved into the math behind volume rendering and neural 3D scene representation, then built NeRF from scratch using PyTorch. Collaboratively trained it to capture a detailed 3D Wolverine model with perfect lighting, pose, and textures. Finally, I optimized rendering parameters to achieve stunningly realistic 3D reconstructions, pushing the boundaries of this cutting-edge technology. |
 |
Built a Seq2Seq Transformer from scratch for English-to-Spanish translation. The decoder employs attention mechanisms to analyze each English word and contextually generate corresponding Spanish phrases. Trained on a curated bilingual dataset, the model utilizes backpropagation and gradient descent to refine its translation accuracy. This self-made language bridge opens doors for further exploration of advanced attention mechanisms, complex sentence structures, and even multilingual translation. |
|
TerraIncognita cracks the code of aerial photos. Trained on the FPN architecture, it maps Earth's land cover, pixel by pixel. From smart urban planning to greener monitoring and sharper farming, this project unlocks a future woven from precise insights. |
|
Implemented object detection using YOLO V3 from scratch. |
This template is a modification to Jon Barron's website and a fork of Hardik Shah's website. Find the source code to my website here.