DebdipCS/Latent-Resonance-AI-Image-Forensics-Benchmark-N100

Latent Resonance: SOTA Empirical AI Image Forensics Benchmark (N=100 & N=1,000 Scale)

Author: Debdip Bandyopadhyay (Independent AI Researcher, Kolkata, India; M.Tech, IIT Jodhpur, AI & Data Science)
Preprint & Paper: Latent Resonance: Zero-Shot Autoencoder Inversion and Azimuthal Spectral Forensics for Diffusion Image Attribution (IEEE Flagship / CERN Zenodo 2026)

---

Benchmark Overview

This repository provides:

1. The official verified $N=100$ ground-truth image dataset (PNG image pairs and metadata).
2. The complete $N=1,000$ large-scale empirical audit results (benchmark_predictions.csv and publication_sota_graphic.png).

For the standalone $N=1,000$ benchmark dataset repository, visit:
👉 DebdipCS/Latent-Resonance-AI-Image-Forensics-Benchmark-N1000

---

Large-Scale Benchmark Results ($N=1,000$)

Forensic Metric	Authentic Optical Cameras ($N=500$)	Generative Latent Diffusion ($N=500$)	Separation Margin ($\Delta$)	Statistical Significance
Reconstruction PSNR	$33.28 \pm 0.96\text{ dB}$	$37.15 \pm 0.60\text{ dB}$	$\mathbf{+3.88\text{ dB}}$	$p = 2.88 \times 10^{-165}$
2D-FFT Harmonic Spike	$1.144 \pm 0.086\times$	$3.655 \pm 0.294\times$	$\mathbf{+2.511\times}$	$p < 10^{-300}$
CMOS PRNU ($\rho_{\text{RGB}}$)	$0.000 \pm 0.001$	$0.981 \pm 0.011$	$\mathbf{+0.981}$	Clean non-overlapping
Effect Size (Cohen's $d$)	\multicolumn{3}{c}{$\mathbf{4.84}$ (Immense Effect Separation)}	Huge separation
Clean AUROC	\multicolumn{3}{c}{$\mathbf{100.00%}$}	Zero Accusation Error
False Accusation Rate (FAR)	\multicolumn{3}{c}{$\mathbf{0.00%}$}	Zero False Positives
Synthetic Recall (TPR)	\multicolumn{3}{c}{$\mathbf{100.00%}$}	$500$ / $500$ Detected

---

Dataset Structure

├── real_photos/             # 50 Authentic Optical Camera PNGs (512x512)
├── ai_synthetic/            # 50 Native Latent Diffusion PNGs (512x512)
├── metadata.jsonl           # N=100 JSONL records with labels and metrics
├── benchmark_predictions.csv # N=1,000 complete item-level empirical audit records
└── publication_sota_graphic.png # 6-panel publication diagnostic graphic

---

Quick Usage with Hugging Face datasets

from datasets import load_dataset

dataset = load_dataset("DebdipCS/Latent-Resonance-AI-Image-Forensics-Benchmark-N100")
print(dataset)

Citation

@article{bandyopadhyay2026latent,
  title={Latent Resonance: Zero-Shot Autoencoder Inversion and Azimuthal Spectral Forensics for Diffusion Image Attribution},
  author={Bandyopadhyay, Debdip},
  journal={CERN Zenodo Open-Access Archive / IEEE Preprint},
  doi={10.5281/zenodo.22158286},
  year={2026}
}