Research Assistant @ MBZUAI · Incoming M.S. Student · B.Tech ECE, VIT Vellore
Building vision-language models for geospatial understanding
I am a researcher working at the intersection of computer vision, vision-language models, and geospatial AI. Currently a Research Assistant at the Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI), where I am building Tera-SAM — an agentic segmentation pipeline for zero-shot instance segmentation of satellite imagery using SAM3 and VLMs.
I completed my B.Tech in Electronics and Communication Engineering from VIT Vellore (GPA: 8.29) and am preparing to begin graduate studies. My work spans the full stack from model design to edge deployment on Raspberry Pi and NVIDIA Jetson hardware.
Current focus → VLMs for geospatial segmentation · Agentic AI · Remote sensing
Past work → IoT security · Video restoration · Edge CV · Assistive tech
| Period | Role | Institution |
|---|---|---|
| Dec 2025 – Present | Research Assistant | Mohamed Bin Zayed University of AI (MBZUAI) |
| May 2025 – Sep 2025 | Research Intern | Indian Institute of Technology, Guwahati |
| Jan 2024 – Aug 2024 | Research Intern | National University of Singapore |
| Jun 2024 – Jul 2024 | Research Intern | Indian Institute of Technology, Dharwad |
| Mar 2024 – Jun 2024 | Computer Vision Intern | Vellore Institute of Technology |
| Sep 2023 – Nov 2023 | Data Science Intern | Quick Heal Technologies |
📄 AIoT Enhanced Night Vision Object Detection for Vehicle Safety — IEEE NETAPPS 2024
📄 Optimizing IoT Security: A ML Pipeline for Fast Intrusion Detection — IEEE NETAPPS 2024
📄 VocalEyes: Vision-Language Models for the Visually Impaired — IEEE ICEI 2024 @ IIT Dharwad
📄 NoiseFed: Federated Learning for Breast Cancer Detection — Preprint 2024
| Project | Description | Status |
|---|---|---|
| TerraSAM | Agentic VLM pipeline for geospatial segmentation (SAM3 + LLM) | 🟢 Ongoing |
| Video Restoration | Weather-degraded video restoration — Swin + Restormer (+5% PSNR) | 🔵 Research |
| DarkRoad IoT Guardian | Night-vision object detection on Raspberry Pi 5 — IEEE Published | ✅ Complete |
| NetSentry IoT | ML-driven network intrusion detection — 0.952 accuracy — IEEE Published | ✅ Complete |
| VocalEyes | Assistive vision system for the visually impaired using Florence-2 on Jetson | ✅ Complete |
Languages
Frameworks & Libraries
Hardware & Platforms
Certifications