AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

PICS: Pairwise Image Compositing with Spatial Interactions

Hang Zhou, Xinxin Zuo, Sen Wang, Li cheng

ICLR 2026 · Image generation › Autoregressive and latent generation · approach: Compositional scene generation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace MLP object-fusion with the interaction transformer block (ablation)PICS · In-the-wild test setFID: 173.1 -> 165.2Improved
replace MLP object-fusion with the interaction transformer block (ablation)PICS · In-the-wild test setCLIP-score: 74.6 -> 76.3Improved
add in-plane rotation augmentation (ablation)PICS · In-the-wild test setFID: 165.2 -> 162.5Improved
add in-plane rotation augmentation (ablation)PICS · In-the-wild test setCLIP-score: 76.3 -> 74.9Hurt
train on combined 1M-image data instead of LVIS only (ablation)PICS · In-the-wild test setFID: 162.5 -> 158.2Improved
train on combined 1M-image data instead of LVIS only (ablation)PICS · In-the-wild test setCLIP-score: 74.9 -> 77.3Improved
add multi-view shape prior (ablation)PICS · In-the-wild test setFID: 158.2 -> 151.3Improved
add multi-view shape prior (ablation)PICS · In-the-wild test setCLIP-score: 77.3 -> 79.1Improved
use PICS parallel compositing instead of AnyDoor sequential compositingPICS · LVIS val (recompositing)FID: 27.17 -> 24.99Improved
use PICS instead of AnyDoorPICS · LVIS val (recompositing)DINOv2-score: 0.6089 -> 0.5713Hurt
use PICS instead of OmniPaintPICS · DreamBoothDreamSim: 0.2632 -> 0.3054Hurt
use PICS instead of ObjectStitchPICS · DreamBoothFID: 260.4 -> 255.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy