AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Cross-Modal Progressive Comprehension for Referring Segmentation

Si Liu, Tianrui Hui, Shaofei Huang, Yunchao Wei, Bo Li, Guanbin Li

TPAMI 2021 · Vision-language and multimodal › Human and embodied modeling · approach: Referring expression comprehension · 156 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 entity perception stage (ablation)baseline · UNC valoverall IoU: 47.36 -> 49.06Improved
add relation-aware reasoning stage (ablation)baseline · UNC valoverall IoU: 47.36 -> 53.40Improved
add EP and RAR together (ablation)baseline · UNC valoverall IoU: 47.36 -> 55.38Improved
add GloVe initialization (ablation)CMPC-I (EP+RAR) · UNC valoverall IoU: 55.38 -> 56.00Improved
add CMF (concatenate multimodal feature) (ablation)CMPC-I with GloVe · UNC valoverall IoU: 56.00 -> 57.12Improved
add TGFE to multi-level baseline (ablation)multi-level baseline · UNC valoverall IoU: 56.38 -> 58.81Improved
increase TGFE exchange rounds from 1 to 3 (ablation)CMPC-I + TGFE · UNC valoverall IoU: 60.72 -> 61.25Improved
add necessary words extraction (ablation)TGFE full model · UNC valoverall IoU: 61.23 -> 61.80Improved
add action-aware reasoning (AAR) (ablation)video model with TGFE, EP, RAR · A2D SentencesOverall IoU: 63.2 -> 64.9Improved
use direct relevance instead of adaptive relevance in AAR (ablation)CMPC-V · A2D SentencesOverall IoU: 64.5 -> 64.9Improved
use 1 instead of 0 graph convolution layers (ablation)CMPC-I · UNC valoverall IoU: 49.06 -> 55.38Improved
use 2 instead of 1 graph convolution layers (ablation)CMPC-I · G-Ref valoverall IoU: 38.19 -> 40.12Improved

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