This repository is based on shouxieai/tensorRT_Pro, with adjustments to support YOLOv8.

Converting COCO annotation (CVAT) to annotation for YOLO-seg (instance segmentation) and YOLO-obb (oriented bounding box detection)

Tensorrt codebase to inference in c++ for all major neural arch using onnx

This powerful and flexible application, developed using PyQt6, processes 4 video streams in real-time with the YOLOv8-OBB model, supporting cameras, video files, for efficient object detection and result saving, tailored to meet specific client requirements.

YOLOv8 Oriented Bounding Boxes tester

End-to-end, reproducible pipeline for fine-tuning YOLOv8-OBB on SAR imagery, including dataset extraction and validation, exploratory data analysis, annotation cleaning, DOTA-to-YOLO OBB conversion, optimized training, rigorous evaluation, and interactive Gradio deployment for accurate oriented vehicle detection on SIVED.

This project is a YOLO based surveillance system designed to assist wildlife conservation efforts. It utilizes a dual-model computer vision pipeline to detect wildlife species and identify potential poaching threats in real-time.

Diamorph object detection