FakeMeh is a web platform that helps users verify images and links by detecting deepfakes. Users can upload an image or URL, and our system analyzes it using the SightEngine API, providing a likelihood percentage of AI manipulation. A heatmap visualization via the MXFace API highlights altered regions, making detection more transparent. The MXFace API plays a pivotal role in your deepfake detection project by providing the foundational facial analysis data that enables heatmapping. Fakemeh also features an interactive game where users guess if an image is real or fake, earning badges and points to promote media literacy.
We developed FakeMeh using React.js for the front end and Flask for the backend, integrating the SightEngine API for deepfake detection. Our database tracks user progress and rewards to enhance engagement.
Optimizing the heatmap for clarity and ensuring efficient API processing were major challenges. Balancing usability and accuracy also required fine-tuning.
We successfully built an engaging deepfake detection tool, combining AI analysis with an educational experience.
We gained expertise in AI-driven image analysis, API integration, gamification strategies, and software engineering.
We plan to detect video deepfakes, introduce leaderboards, and integrate the Google Fact Check API for better source verification.
Built With
- googlefactchecker
- javascript
- mxface
- node.js
- npm
- python
- react
- sightengine
Log in or sign up for Devpost to join the conversation.