Completed
SubseaVision - Underwater Object Detection
Machine learning pipeline for detecting and classifying underwater objects, marine life, and infrastructure defects from ROV video feeds.
Status
completed
Technologies Used
PyTorch OpenCV TensorRT CUDA FastAPI
About This Project
SubseaVision uses state-of-the-art deep learning models to automatically detect and classify objects in underwater imagery. The system can identify pipeline damage, marine growth, anode degradation, and various marine species with high accuracy.