AI Paper Atlas

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

Fine-Grained Video-Text Retrieval With Hierarchical Graph Reasoning

Shizhe Chen, Yida Zhao, Qin Jin, Qi Wu

CVPR 2020 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 332 citations (OpenAlex)

Check 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
decompose matching into event, action, and entity levels and fuse their scores (ablation)HGR · MSR-VTTrsum: 57.6 -> 71.9Improved
add graph attention instead of averaging neighboring nodes (ablation)HGR · MSR-VTT text-to-videoR@10: 35.6 -> 36.5Improved
add graph attention instead of averaging neighboring nodes (ablation)HGR · MSR-VTT video-to-textR@10: 47.1 -> 48.8Improved
add semantic-role awareness to graph reasoning (ablation)HGR · MSR-VTT text-to-videoR@1: 9.1 -> 9.2Improved
add semantic-role awareness to graph reasoning (ablation)HGR · MSR-VTT video-to-textR@1: 14.2 -> 15.0Improved
add hierarchical video embeddings (ablation)HGR · MSR-VTTrsum: 167.9 -> 172.4Improved
use HGR instead of Dual Encodingvideo-text retrieval · MSR-VTT testrsum: 148.6 -> 172.4Improved
use HGR instead of PVSEvideo-text retrieval · TGIF text-to-videoR@1: 2.3 -> 4.5Improved
use HGR instead of VSE++video-text retrieval · VATEX text-to-videoR@1: 33.7 -> 35.1Improved
train on MSR-VTT and evaluate HGR on an unseen datasetHGR · YouTube2Text text-to-videoR@1: 13.8 -> 16.4Improved
use HGR rather than Dual Encoding for switching-role discriminationfine-grained binary selection · YouTube2Text switch rolesaccuracy: 71.92 -> 69.48Hurt
use HGR rather than VSE++ for incomplete-event discriminationfine-grained binary selection · YouTube2Text incomplete eventsaccuracy: 78.79 -> 82.04Improved

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