An interactive tool to visualize and understand what deep learning models "see" using feature maps and Grad-CAM.
NeuroLens helps interpret deep learning models by visualizing internal activations and highlighting important regions in images. It makes neural networks more explainable and transparent.
- Feature map visualization
- Grad-CAM heatmaps
- Support for pretrained models (ResNet, VGG)
- Interactive UI (React + FastAPI)
- Python (PyTorch, OpenCV)
- FastAPI (backend)
- React (frontend)
- git clone https://github.com/Ash3010/NeuroLens.git
- cd NeuroLens
- pip install -r requirements.txt
streamlit run app.py