DeepFake Detector
Developed the user interface and helped with integration for a deepfake detection project with Purdue University Fort Wayne, USA.
My focusUI development & integration support
The public HTML/CSS/JavaScript prototype uses simulated API responses. It demonstrates the interface and interaction design.
Overview
A deepfake detection project with Purdue University Fort Wayne, USA. My primary contribution was developing the user interface, with additional support for application integration.
The problem
Manipulated facial images can be difficult to identify by visual inspection alone. The project explores model-based detection and a backend interface that applications can integrate.
Approach
Developed a React dashboard with interactive probability visualizations, heatmap overlays, and frame-by-frame diagnostic views. These interfaces make the detection model's outputs easier to inspect. Also supported integration with the broader application.
System design
The broader system combines a detection model, image processing, and a backend service. My primary responsibility was the React dashboard and supporting application integration, rather than ownership of the entire model or backend.
Results & evaluation
Delivered the dashboard and supported application integration for the project with Purdue University Fort Wayne. The interface presents model probabilities and visual diagnostics so users can inspect results.