Benchmark Radar
AI BENCHMARK PROFILE

ClueAegis-Bench

General AIMultimodal Perception

A benchmark decomposing synthetic image detection into annotated forensic cognitive skills, used to evaluate a proposed detection framework.

Released
2026-05-24
Readiness
Paper only
Primary field
General AI

Why it matters

Supports evidence-based synthetic image detection beyond binary classification.

Motivation

The rapid advancement of generative models has made synthetic images increasingly realistic, challenging reliable detection.

Primary resources

Benchmark Radar records only publicly supported details and links back to primary sources for verification.