Connect the parts
Input Image -> OpenCV Noise Residual Preprocessing -> Vision Transformer (ViT) Feature Embeddings -> Binary Classification Head (Real vs Generated).
COMPUTER VISION / archived
Real vs AI Image Detector
A deep-learning powered system that identifies whether an image is AI-generated or real using advanced image forensics and transformer-based feature extraction..
The problem.
Explore whether visual artifacts help distinguish generated and real images.
My contribution.
Designed and implemented the application described in this project, including its core workflow and supporting interface.
How the parts connect.
Input Image
Part of the project architecture.
OpenCV Noise Residual Preprocessing
Part of the project architecture.
Vision Transformer (ViT) Feature Embeddings
Part of the project architecture.
Binary Classification Head (Real vs Generated).
Part of the project architecture.
The decisions.
Keep the workflow focused
AI-vs-Real image classification
Build, test, deliver.
Implementation and setup instructions.
What to keep in mind.
This is a portfolio study. No independently verified performance or accuracy benchmark is claimed.
Repository availability and external demonstrations can change; consult the source for current setup and limitations.