AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Rethinking Visual Rearrangement from A Diffusion Perspective

Tianliang Qi, Xinhang Song, Yuyi Liu, Shuqiang Jiang

CVPR 2026 · Image generation › Fast diffusion sampling · approach: Diffusion theory and design space

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
use diffusion-bridge denoising transformer instead of Kuhn-Munkres matching (ablation)Diffusion Rearrangement · RoomR testSuccess (%): 12.8 -> 17.8Improved
use denoising transformer instead of direct matching (ablation)Diffusion Rearrangement · RoomR testSuccess (%): 6.5 -> 17.8Improved
represent objects as Gaussian mixture distributions instead of center coordinates (ablation)Diffusion Rearrangement · RoomR testSuccess (%): 16.7 -> 17.8Improved
represent objects as Gaussian mixture distributions instead of PointNet++ features (ablation)Diffusion Rearrangement · RoomR testSuccess (%): 17.2 -> 17.8Improved
represent objects as GMM instead of coordinates (ablation)Diffusion Rearrangement · RoomR testFixed Strict (%): 37.1 -> 38.8Improved
increase sampling steps from 20 to 200 (ablation)Diffusion Rearrangement · RoomR validationSuccess (%): 16.4 -> 17.9Improved
increase sampling steps beyond 200 (ablation)Diffusion Rearrangement · RoomR validationSuccess (%): 17.9 -> 17.4 (1000 steps)Hurt
add Gaussian noise to depth input (ablation)Diffusion Rearrangement · RoomR testSuccess (%): 17.8 -> 17.5Hurt
diffusion-based inference instead of matching-based change inferenceDiffusion Rearrangement vs CAVR · RoomR testSuccess (%): 14.2 -> 17.8Improved
diffusion-based inference instead of matching-based change inferenceDiffusion Rearrangement vs CAVR · ProcTHOR (self-built) testSuccess (%): 4.9 -> 8.4Improved

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