// case study
YOLO garbage detection
A YOLO model that detects and classifies dozens of litter types in field photos.
Context
An artificial intelligence project from my previous portfolio.
What I built
- An object detection model based on the YOLO architecture: convolution and max-pool layers followed by fully connected layers.
- Training on dozens of litter classes, e.g. plastic film, clear plastic bottle, plastic bottle cap, drink can, corrugated carton, and single-use carrier bag.
- Evaluation with a per-class confusion matrix, then testing on field photos with bounding boxes and confidence scores.