Image Classifier — MobileNetV2 + Grad-CAM
A simple, lightweight image classification project built using MobileNetV2, fine-tuned on 5 flower categories, with Grad-CAM explainability. Upload an image → get a prediction → see a heatmap showing what the model focused on.
Built entirely on a simple laptop, beginner-friendly, and fully open-source.
-MobileNetV2 image classifier (fast + lightweight)
-5 flower classes: daisy, dandelion, rose, sunflower, tulip
-Grad-CAM heatmaps for explainability
-Clean and simple Gradio web interface
-CPU-friendly (no GPU required)
-Beginner-friendly project structure
-Python
-PyTorch
-Torchvision
-Gradio
-NumPy
-Matplotlib
This was built to understand:
-Transfer Learning
-Image Classification
-Grad-CAM Explainability
-Building AI apps on a low-end laptop
-Deploying simple ML demos