Built a ML model using YOLOv8 to estimate the number or survivors in high-risk buildings by counting people entering and leaving through CCTV camera's.
- Yolov8 model
- Flask
- Codepen
THe model achieved a accuracy of 87.2%.
Annotated over 1750+ images to create a dataset by Roboflow from raw videos and developed a video upload interface and streamlined with flask backend operations.
