PROVE: A Perceptual RemOVal cohErence Benchmark for Visual Media
Fuhao Li, Shaofeng You, Jiagao Hu, Yu Liu, Yuxuan Chen, Zepeng Wang, Fei Wang, Daiguo Zhou, Jian Luan

TL;DR
PROVE introduces perception-aligned metrics RC-S and RC-T, along with PROVE-Bench, a comprehensive benchmark, to improve evaluation of object removal in images and videos by better aligning with human perception.
Contribution
The paper presents RC metrics and PROVE-Bench, addressing limitations of existing metrics and establishing a new evaluation framework for visual media object removal.
Findings
RC metrics outperform existing metrics in aligning with human judgments.
PROVE-Bench provides a challenging real-world dataset for benchmarking.
Experiments validate the effectiveness of RC metrics across diverse benchmarks.
Abstract
Evaluating object removal in images and videos remains challenging because the task is inherently one-to-many, yet existing metrics frequently disagree with human perception. Full-reference metrics reward copy-paste behaviors over genuine erasure; no-reference metrics suffer from systematic biases such as favoring blurry results; and global temporal metrics are insensitive to localized artifacts within edited regions. To address these limitations, we propose RC (Removal Coherence), a pair of perception-aligned metrics: RC-S, which measures spatial coherence via sliding-window feature comparison between masked and background regions, and RC-T, which measures temporal consistency via distribution tracking within shared restored regions across adjacent frames. To validate RC and support community benchmarking, we further introduce PROVE-Bench, a two-tier real-world benchmark comprising…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
