AI Automation
TB Detection AI
A computer vision model that classifies Tuberculosis from chest X-ray images, built with transfer learning. A full pipeline: dataset collection, augmentation, training with class imbalance handling, evaluation (confusion matrix & ROC-AUC), Grad-CAM visualization to show where the model focuses, and inference on user-uploaded X-rays. Note: an educational/research project, not a medical diagnostic tool.
Key Features
- Normal vs Tuberculosis classification from X-rays
- Transfer learning + fine-tuning
- Augmentation & class weighting for imbalanced data
- Evaluation: confusion matrix, ROC-AUC, precision/recall
- Grad-CAM: visualizing model focus areas
- Direct image upload inference
TB Detection AIComputer VisionChest X-ray Classification · Grad-CAM
Need a similar system for your business?
Free Consultation