> SELECTED WORK.
> CASE STUDIES.
Built an ML pipeline to predict PM2.5 using satellite AOD, MERRA-2, INSAT-3D, and CPCB datasets with Streamlit dashboards.
Developed dual ESP32-CAM stereo vision depth estimation with YOLOv8 and TensorFlow for real-time robotic/UAV perception.
Built an autonomous embedded robotics system with ultrasonic sensing, laptop processing, and obstacle-navigation logic.
Designed a sensor-fusion architecture for an autonomous owner-following trolley using vision tracking and localization concepts.
Researching Liquid Neural Networks (LNNs) integrated with Transformer architectures for adaptive edge AI systems. Exploring real-time perception systems, efficient AI inference, sequential learning workflows, and autonomous UAV navigation for aerial robotics.