Due to particular law regulation, we are unable to showcase our project
But here are the video we won 1st place out of 200 teams during the annual 2023 Captone Competition
https://youtu.be/ABhfHQkjvkU?si=wspO8OWele_voZWu&t=10152
We designed an advanced AI project aimed at optimising trash bag collection processes for NetVision Company . Leveraging the power of YOLOv8, our team created a sophisticated system capable of detecting, classifying, segmenting, tracking, and estimating the size of trash bags within real-time video streams. My part included reading research papers and training models on YOLOv 8. With this project, we participated in the capstone competition 2023 at Woosong University. We defeated more than 200 teams from 29 departments and got the Woosong University President Award. Here is the video of the awarding our team : Video Link . For security purposes, I will not give access to the project, but tell briefly in the following paragraphs: Technical Strategy: YOLOv8 Integration: Employed YOLOv8, a state-of-the-art object detection model, to effectively identify and categorise trash bags within live video streams. Real-time Processing:Engineered algorithms for immediate analysis of video data to swiftly and accurately identify and track trash bags. Precise Segmentation: Utilised advanced segmentation techniques to precisely isolate trash bags in video frames for detailed analysis. Project Deliverables: Detection & Classification: Accurate identification and classification of various types of trash bags (e.g, recyclable, non-recyclable) within video feeds. Segmentation & Tracking: Precise isolation and continuous tracking of trash bags throughout the video streams for efficient monitoring. Size Estimation: Providing estimations of trash bag sizes for optimal collection planning and resource allocation. Tools Used: PyTorch, OpenCV, YOLOv8, Ultralytics ML Algorithms: YOLOv8 object detection