Common Inpainted Objects In-N-Out of Context

Tianze Yang*, Tyson Jordan*, Ruitong Sun*, Ninghao Liu, Jin Sun
University of Georgia
*Equal Contribution
CVPR 2026

Abstract

We present Common Inpainted Objects In-N-Out of Context (COinCO), a novel dataset addressing the scarcity of out-of-context examples in existing vision datasets. By systematically replacing objects in COCO images through diffusion-based inpainting, we create 97,722 unique images featuring both contextually coherent and inconsistent scenes, enabling effective context learning. Each inpainted object is meticulously verified and categorized as in- or out-of-context through Large Vision Language Model assessments. Our analysis reveals significant patterns in semantic priors that influence inpainting success across object categories. We demonstrate three key tasks enabled by COinCO: (1) developing a fine-grained context reasoning approach that classifies objects as in- or out-of-context based on three criteria; (2) a novel Objects-from-Context prediction task that determines which new objects naturally belong in given scenes at both instance and clique levels, and (3) context-enhanced fake detection on state-of-the-art methods without fine-tuning. COinCO provides a controlled testbed with contextual variations, establishing a foundation for advancing context-aware visual understanding in computer vision and image forensics.

Overview

COinCO Pipeline Overview

Overview of the COinCO dataset construction pipeline and downstream tasks.

Example Images from COinCO

Explore More

Dive deeper into our dataset construction, analysis, and downstream tasks.

BibTeX

@inproceedings{yang2026coinco,
  title={Common Inpainted Objects In-N-Out of Context},
  author={Yang, Tianze and Jordan, Tyson and Sun, Ruitong and Liu, Ninghao and Sun, Jin},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2026},
  eprint={2506.00721},
  archivePrefix={arXiv}
}