Camera-calibrated computer vision system for real-world object detection, segmentation, and metric dimension estimation using Mask R-CNN.

AI-powered PPE compliance and restricted zone monitoring for construction sites. YOLOv8 object detection + DeepSORT tracking with automatic violation logging and snapshots.

Multi-class traffic sign classification using TensorFlow/Keras CNN. Custom preprocessing , accuracy curves, classification report and confidence-score predictions.

Real-time restricted area surveillance using OpenCV. Manual ROI selection, MOG2 background subtraction, motion alerts and automatics timestamp logging. Beginner-friendly.

Beginner-friendly image classification using PyTorch and ResNet-18 transfer learning. Clean workflow, easy to follow for students.