AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Weakly Supervised Referring Image Segmentation with Intra-Chunk and Inter-Chunk Consistency

Jungbeom Lee, Sungjin Lee, Jinseok Nam, Seunghak Yu, Jaeyoung Do, Tara Taghavi

ICCV 2023 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Weakly supervised semantic segmentation · 21 citations (OpenAlex)

Read the paper

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 intra-chunk and inter-chunk consistency losses (ablation)ALBEF Grad-CAM · RefCOCO+ valmIoU: 22.44 -> 25.68Improved
add patch-affinity refinement from ViT self-attention (ablation)ALBEF with chunk consistency · RefCOCO+ valmIoU: 25.68 -> 29.58Improved
add MCG unsupervised shape prior (ablation)ALBEF with chunk consistency and affinity · RefCOCO+ valmIoU: 29.58 -> 31.28Improved
add MCG shape prior (ablation)Ours · RefCOCO+ valPr@70: 1.83 -> 9.72Improved
add intra-chunk and inter-chunk consistency (ablation)ALBEF · RefCOCO+ valPr@50 and Pr@70: Pr@70 0.20 -> 0.34No clear effect
use layer 10 instead of layer 9 for patch affinity (ablation)affinity refinement · RefCOCO+ valmIoU: 29.58 -> 20.71Hurt
exclude negative captions from intra/inter-chunk losses (ablation)Ours · RefCOCO+ valmIoU: 25.68 -> 25.76No clear effect
add box supervision loss via BBAM pseudo maskOurs · G-Ref valoIoU: 46.03 -> 49.64 vs Feng et al.Improved
add chunk consistency to ALBEF for inclusive relation captionsALBEF · RefCOCO+ val subset, inclusivemIoU: 24.46 -> 28.72Improved
add chunk consistency to ALBEF for exclusive relation captionsALBEF · RefCOCO+ val subset, exclusivemIoU: 19.62 -> 23.57Improved

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