Backed by advanced LLM Vision APIs. Powered by Generative AI.
Delivering unparalleled insight for currency verification.
BUILT BY RAUSHAN KUMAR · NIT NAGALAND
A system that takes a currency note image as input, extracts security features, compares them against authentic references using SSIM, and delivers a detailed pass/fail verdict.
Pixel-level extraction and analysis of security features from currency note images.
Individual verification of multiple security features against authentic reference samples.
Structural Similarity Index provides quantitative scores for each security feature.
Instead of hardcoded rules, the system intelligently reasons through its analysis with deep learning.
Upload an image or use your camera to analyze currency authenticity in real-time.
Drop currency image here or browse
JPG, PNG, WEBP up to 10MB
Upload an image or use camera to begin analysis
Evaluates visual consistency and micro-structures intelligently using Gemini 1.5 Pro Vision.
Serverless architecture directly communicating visual buffers to deep-learning endpoints.
Presents human-readable explanations instead of basic pass/fail binary limits.
Generates real-time authenticity probabilities based on LLM heuristic interpretations.
Images are uploaded directly to the provider. No intermediate backend logging or storage.
Capable of learning and assessing any international currency via zero-shot classification.
Input your secure Gemini API Key via the settings panel at the top right.
Use your local webcam or browse for a high-quality photo of the currency.
The system converts the image pixel data directly to Base64 and transmits to Google servers.
Gemini LLM processes visual security features utilizing multi-modal deep-learning.
Detailed, human-readable structural analysis and boolean authenticity are displayed.
