AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

CObL: Toward Zero-Shot Ordinal Layering without User Prompting

Aneel Damaraju, Dean Hazineh, Todd Zickler

ICCV 2025 · Detection and segmentation › Open-set and few-shot detection · approach: Zero-shot segmentation and detection · 1 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
add MiDaS depth cue to the conditioning adapter (ablation)CObL · TableTop ablationLPIPS: LPIPS increases by 4% without itImproved
keep Stable Diffusion UNet frozen instead of unfreezing it (ablation)CObL · TableTopLPIPS: 9% increase in LPIPS when unfrozenImproved
add prior score matching and compositional guidance at inference (ablation)CObL · TableTopLPIPS: 8% LPIPS increase without guidanceImproved
increase number of objects in scene beyond four (ablation)CObL · TableTopLPIPS and CLIPHurt
add non-differentiable permute, erase and sort updates during sampling (ablation)CObL · TableTopevaluation metricsNo clear effect
use CObL instead of pix2gestalt with oracle masksamodal completion · TableTop (60 scenes, 2-4 objects)LPIPS best: .128 -> .094Improved
use CObL instead of pix2gestalt with oracle masksamodal completion · TableTop (60 scenes, 2-4 objects)CLIP best: .889 -> .935Improved
use CObL instead of SDXL Inpainting with oracle masksamodal completion · TableTop (60 scenes, 2-4 objects)LPIPS best: .373 -> .094Improved
use CObL instead of LaMa with oracle masksamodal completion · TableTop (60 scenes, 2-4 objects)LPIPS best: .113 -> .094Improved
use CObL instead of pix2gestaltamodal completion · TableTop (60 scenes, 2-4 objects)LPIPS average: .153 -> .122Improved
use CObL instead of Mask2Former for visible segmentationpanoptic segmentation · TableTop visible segmentationARI: 66.3% -> 83.5%Improved

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