Deep learning framework for image authenticity prediction with three core experiments: (1) CNN architecture comparison & network pruning analysis, (2) Model explainability comparison (GradCAM vs Multiscale Pixel Masking), (3) Ensemble learning strategies for authenticity detection.
Cryptographic image authenticity verification — prove edits are localized without exposing the original. Fully client-side.
Hardware-backed image authenticity and watermarking system for secure, compression-resilient provenance.
Cryptographic image integrity & authenticity verification tool. Detects any image modification via pixel-level validation.
AI-powered image authenticity analyzer built with React, TypeScript, Vite, Tailwind CSS, and Google Gemini. Detect potential AI-generated or digitally manipulated images with explainable visual analysis.
To associate your repository with the image-authenticity topic, visit your repo's landing page and select "manage topics."