TruthLens AI is an advanced multi-modal authenticity detection framework developed to evaluate digital content trustworthiness in both Marathi and English languages. It supports multiple formats including text news, social media links, image screenshots, and video metadata.
Created by Shrutika Sonmale as a capstone digital trust solution to combat regional misinformation, edited screenshots, viral fake claims, and AI-manipulated content.
Analyzes news articles and WhatsApp forwards in Marathi and English to detect clickbait patterns, deceptive headlines, and misleading claims.
Integrated with Google Fact-Check Tools API to instantly match claims against verified databases from trusted fact-checking organizations.
Combines Tesseract OCR (Marathi + English) with Error Level Analysis (ELA) to extract embedded text and detect Photoshop or pixel manipulations in screenshots.
Performs temporal frame variance and inconsistency checks using OpenCV to flag deepfakes and manipulated video files.