AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Multi-Modal Relational Graph for Cross-Modal Video Moment Retrieval

Yawen Zeng, Da Cao, Xiaochi Wei, Meng Liu, Zhou Zhao, Zheng Qin

CVPR 2021 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 83 citations (OpenAlex)

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 GCN feature optimization to the Backbone (ablation)Charades-STA retrieval model · Charades-STAR@1, IoU=0.3: 64.38% -> 64.37%Hurt
Add GAT feature optimization to the Backbone (ablation)Charades-STA retrieval model · Charades-STAR@1, IoU=0.3: 64.38% -> 66.72%Improved
Add cross-graph attention to visual and textual graphs (ablation)Charades-STA retrieval model · Charades-STAR@1, IoU=0.3: 66.72% -> 66.35%Hurt
Add spatio-temporal graph modeling (ablation)Charades-STA retrieval model · Charades-STAR@1, IoU=0.3: 64.38% -> 67.73%Improved
Add explicit interaction-relation attributes (ablation)+STG · Charades-STAR@1, IoU=0.3: 67.73% -> 68.27%Improved
Add attribute-masking pre-training (ablation)+RSTG · Charades-STAR@1, IoU=0.3: 68.27% -> 70.08%Improved
Add context-prediction pre-training (ablation)+RSTG · Charades-STAR@1, IoU=0.3: 68.27% -> 69.73%Improved
Combine attribute masking and context prediction (ablation)+RSTG · Charades-STAR@1, IoU=0.3: 68.27% -> 71.60%Improved
Use multi-modal relational graphs, graph pre-training, graph matching, and boundary regressionCross-modal video moment retrieval · Charades-STAR@1, IoU=0.3: 70.55% -> 71.60%Improved
Use multi-modal relational graphs, graph pre-training, graph matching, and boundary regressionCross-modal video moment retrieval · TACoSR@1, IoU=0.3: 57.53% -> 57.83%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