AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Contrastive Region Guidance: Improving Grounding in Vision-Language Models without Training

David Wan, Jaemin Cho, Elias Stengel-Eskin, Mohit Bansal

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 45 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 contrastive region guidance (blackout contrast) at decodingLLaVA-1.6-34B · ViP-Benchaverage score: 45.0 -> 56.1Improved
add CRG on top of a visual-prompt fine-tuned modelViP-LLaVA-13B · ViP-Benchaverage score: 48.6 -> 49.4Improved
add Set-of-Mark overlay prompting to open-source VLMsLLaVA-1.6-34B and ViP-LLaVA-13B · ViP-Benchaverage score: 45.0 -> 42.0Hurt
overlay bounding boxes without CRG (ablation)LLaVA-1.6-34B · What'sUp Set of 4accuracy: 54.0 -> 39.7Hurt
add CRGLLaVA-1.6-34B · What'sUp Set of 4accuracy: 54.0 -> 64.0Improved
add CRGLLaVA-1.6-34B · SugarCrepe swap-objaccuracy: 76.3 -> 87.8Improved
add CRGLLaVA-1.6-34B · SeeTRUE averageAUROC: 65.7 -> 74.1Improved
add CRG re-ranking of detector proposalsGroundingDINO-B + LLaVA-1.5-13B · RefCOCO testA (multiple proposals only)accuracy@0.5: 62.0 -> 66.9Improved
add CRG re-rankingGroundingDINO-B + LLaVA-1.5-13B · RefCOCO+ testB (full)accuracy@0.5: 60.4 -> 60.0No clear effect
black out each object separately instead of combined mask (ablation)CRG · What'sUp Set of 4, LLaVA-1.6-34Baccuracy: 60.1 -> 64.0Improved
black out entire image instead of only relevant objects (ablation)CRG · What'sUp Set of 4, LLaVA-1.6-34Baccuracy: 64.0 -> 53.8Hurt
use segmentation-mask blackout instead of bounding-box blackout (ablation)CRG · What'sUp Set of 4, LLaVA-1.6-34Baccuracy: 64.0 -> 55.7Hurt

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