Media Forensics

AI image detection using signal processing and statistical analysis

Drop an image here or click to browse

JPEG, PNG, or WebP -- up to 10 MB

How It Works

1. Upload

Drop or select any JPEG, PNG, or WebP image up to 10 MB.

2. Analyze

9-step forensic pipeline: ELA, FFT, noise patterns, metadata, and more.

3. Results

Get a detailed report with visualizations, scores, and a final verdict.

9-Step Analysis Pipeline

Metadata Analysis

EXIF data, AI tool signatures, camera info

Error Level Analysis

JPEG re-compression difference patterns

FFT Spectral

Frequency domain patterns and GAN fingerprints

JPEG Ghosts

Multi-quality compression artifact detection

Noise Analysis

Sensor noise patterns and uniformity

Color Statistics

Channel correlations and histogram analysis

Edge Coherence

Edge consistency and sharpness variation

Thumbnail Check

Embedded thumbnail vs main image comparison

Watermark Detection

Hidden markers, LSB analysis, C2PA credentials

Your images are analyzed server-side and stored temporarily for report retrieval. No AI models or external APIs are used -- analysis is performed using signal processing and statistical methods only.