SAM

Object Detection and Mask Prediction System<br> This project leverages the cutting-edge capabilities of YOLOv8 for object detection and incorporates the Segment Anything Model (SAM) as a foundational model for segmentation tasks. This combination is geared towards creating a robust system for detecting objects and accurately segmenting them in various contexts, such as surveillance, medical imaging, and autonomous vehicles.<br>

Features<br>

• YOLOv8 Object Detection: Utilizes the latest iteration of the YOLO (You Only Look Once) model, YOLOv8, known for its efficiency and accuracy in real-time object detection across various contexts.<br>
• Segment Anything Model (SAM): A versatile foundation model used for generating precise segmentation masks for each detected object, enhancing the application's utility in complex segmentation tasks.<be>

Installation<br> Prerequisites:<br>

• Python 3.8+<br>

Usage<br> To operate the system, execute the main script, upload your image, and view the results:<br>

1. Run the main script:<br> python app.py<br>

2. Upload an image:<br>

• Upon execution, the system will prompt you to upload an image.<br>
• The image is processed to detect objects and predict segmentation masks.<br>

1. View results:<br>

• Detected objects and their corresponding masks are stored in a directory.<br>
• Results (masks) are downloaded in a ZIP file.<br>

How It Works<br> YOLOv8 Object Detection<br> YOLOv8, one of the versions in the YOLO series, is employed for its exceptional ability to perform real-time object detection.<br>

Segment Anything Model (SAM)<br> SAM, the Segment Anything Model, serves as the entity for segmentation in this system. After objects are detected using YOLOv8, SAM generates segmentation masks for each detected object.<br>