Benchmark Radar
AI BENCHMARK PROFILE

OCCUR-Bench

General AIMultimodal Perception

OCCUR-Bench evaluates temporal preservation in conversational image editing, providing occlusion-and-revelation scenarios with historical restoration references to assess faithful restoration of occluded content.

Released
2026-07-08
Readiness
Paper only
Primary field
General AI

Why it matters

Targets the under-explored problem of preserving content that temporarily disappears during multi-turn editing, offering a diagnostic benchmark for faithful restoration.

Motivation

Conversational image editing requires preserving not only visible content, but also content that temporarily disappears across turns.

Primary resources

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