https://github.com/adm-2005/demorph
Deepfake Detection Solution using Multimodal Approach.
https://github.com/adm-2005/demorph
deepfake-detection multimodal-deep-learning
Last synced: about 1 month ago
JSON representation
Deepfake Detection Solution using Multimodal Approach.
- Host: GitHub
- URL: https://github.com/adm-2005/demorph
- Owner: Adm-2005
- Created: 2024-12-10T08:38:51.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2024-12-10T08:51:55.000Z (5 months ago)
- Last Synced: 2025-01-27T18:50:09.942Z (3 months ago)
- Topics: deepfake-detection, multimodal-deep-learning
- Language: Python
- Homepage:
- Size: 10.9 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# DeMorph
A comprehensive deepfake detection solution.
## Features
- **Robust Detection:** Analyzes facial movements, lip synchronization, and audio-visual consistency for comprehensive detection.
- **Explainable AI (XAI) Insights:** Provides detailed visual explanations of detected abnormalities using Grad-CAM techniques.
- **Video Processing & Reporting:** Supports real-time and batch processing of videos, ensuring efficiency in deepfake detection. Generates downloadable reports that include authenticity status, detected abnormalities, and confidence scores.
- **Web Content Monitoring:** Lists top websites known for distributing deepfake content, warning users of potential risks.
- **Twitter Trend Analysis:** Allows users to check whether media files uploaded in a twitter trend are authentic or not.
## Technologies Used
- **Languages**: Python, TypeScript
- **Frameworks**: Flask, React, TailwindCSS
- **DL & CV**: PyTorch, OpenCV, GradCAM
- **Database**: PostgreSQL
## Snapshots
![Snapshot1]()
![Snapshot2]()
## Important Links
- GitHub Repo: [Click Here]('https://github.com/Adm-2005/DeMorph-SIH-2024')
- Live Project: [Click Here]('')
- Documentation: [Click Here]('')