dschechter27/vision_model_interpretability

Vision Model Interpretability with Grad-CAM

David Schechter

Upload an image to:

• —classify it using ResNet-18
• —view the top-3 predictions
• —visualize a Grad-CAM heatmap highlighting the image regions that influenced the model’s decision

This demo explores model interpretability in computer vision by showing how gradients from a convolutional neural network can be used to explain predictions.