A passionate Researcher from IIIT-B 📍
- 🔭 Currently working on: Submitted Journal to IJCV on graph based robust VO with epipolar supervision, Bounded NeRF using trajectories from VSLAM/SFM and images, building a simple VO module in C++ and using an open source dataset to implement VLA in kaggle.
- 🌱 Currently learning: Robot Foundation models for embodied AI, Prediction in latent space (V-JEPA), Event VIO and state space estimation.
- 👯 Seeking for research collaborations on: Underwater Image understanding, Multimodal Visual SLAM, 3DGS and VLA.
- 🤔 Looking for help with: Integrating OpenCV C++ libraries with Python Visual SLAM modules for improved 3D reconstruction.
- 🎓 Recent Milestone: Completed Master's thesis and Defence on "GRAPH-BASED RELATIVE POSE ESTIMATION WITH EPIPOLAR GEOMETRY SUPERVISION FOR VISUAL SLAM" at IIIT Bangalore.
- 💬 Ask me about: 3D Computer Vision, Robotics, SLAM, GIS, and Embedded AI.
- 📫 How to reach me: prateeth.rao@iiitb.ac.in | prateeth2012@gmail.com
- ⚡ Fun fact: I love deep-diving into a topic before drawing conclusions, love listening to EDM, bollywood and classical music, and enjoy unwinding with yoga, gardening, or learning about historical facts/structures.
(Click on any project to view the repository and code)
- 🌐 Graph-Based Visual Odometry Codebase supporting my Pattern Recognition submission and Master's thesis, focusing on relative pose estimation with epipolar geometry supervision.
- 📷 Bounded NeRF Implementation Implementation utilizing trajectories from VSLAM/SFM and image data for bounded Neural Radiance Fields.
- 🌊 Coral Reef Mapping - Netrani Island An environmental GIS pipeline leveraging multi-source satellite data (Sentinel-1/2, MODIS, ICESat-2) for underwater analysis.
- ⚙️ C++ Visual Odometry Module A lightweight, from-scratch VO pipeline built in C++ to integrate with broader SLAM architectures.
Here are some of the tools I work with:
Thanks for visiting!