Privacy Policy

ConvLens is an open-source educational tool. We don't track you, we don't show ads, and we don't sell anything. Images you upload are processed in memory to generate visualizations and are not retained beyond the duration of your session.

• Uploaded images. Images you upload for analysis are sent to our backend, processed by a PyTorch model, and used to generate Grad-CAM and feature map visualizations. They are stored temporarily on the server's filesystem during processing and deleted when the server restarts (typically within hours).
• Generated visualizations. Output images (heatmaps, overlays, feature maps) are stored alongside your job temporarily and served back to your browser. These are also ephemeral.
• No personal information. ConvLens does not require accounts, does not collect names, emails, or any identifying information.

• No tracking cookies, no analytics scripts, no ad networks.
• No selling, sharing, or licensing of any data to third parties.
• No long-term storage of uploaded content.
• No model training on user-uploaded images.

Our hosting provider (Railway) maintains standard server logs containing IP addresses and request paths for security and debugging purposes. These are retained per Railway's own policy and are not accessible to ConvLens beyond standard log inspection during incident response.

ConvLens is open source under the MIT license. You can review exactly how your data is handled by reading the source code at github.com/AhmedKamal-41/cnn-layer-visualizer. If you prefer full control, you can self-host the entire stack with onedocker compose upcommand.

For privacy questions, open an issue at github.com/AhmedKamal-41/cnn-layer-visualizer/issues.