AI Paper Atlas

Research map / Video / Video-language understanding

OVG-HQ: Online Video Grounding with Hybrid-modal Queries

Runhao Zeng, Jiaqi Mao, Minghao Lai, Minh Hieu Phan, Yanjie Dong, Wei Wang, Qi Chen, Xiping Hu

ICCV 2025 · Video › Video-language understanding · approach: Temporal video grounding · 7 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
use parametric memory f_PML instead of self-attention layer (ablation)OVG-HQ-Unify · QVHighlights-Unify (text query)oR@1, IoU=0.5: 13.93 -> 23.37Improved
use parametric memory f_PML instead of LSTM (ablation)OVG-HQ-Unify · QVHighlights-Unify (text query)oR@1, IoU=0.5: 22.37 -> 23.37Improved
use f_PML instead of LSTM (ablation)OVG-HQ-Unify · QVHighlights-Unify (Segment-G query)oR@1, IoU=0.5: 17.41 -> 20.13Improved
add anchor features to prediction refinement input (ablation)OVG-HQ-Unify · QVHighlights-Unify (text query)oR@1, IoU=0.5: 18.99 -> 23.37Improved
add prediction refinement module (ablation)OVG-HQ-Unify · QVHighlights-Unify (text query)oR@1, IoU=0.5: 17.64 -> 23.37Improved
update f_PML parameters at inference (Tune vs Frozen) (ablation)OVG-HQ-Unify · QVHighlights-Unify (eight query types)oR@1, IoU=0.5Improved
add hybrid distillation to unified model (Image-R query only) (ablation)OVG-HQ-Unify · QVHighlights-Unify (Image-R query)oR@1, IoU=0.5: 11.43 -> 20.41Improved
train one unified model directly on all query types instead of expert model (Segment-G) (ablation)OVG-HQ-Unify · QVHighlights-Unify (Segment-G query)oR@1, IoU=0.5: 20.33 -> 14.07Hurt
add text to Segment-G query (ablation)expert model · QVHighlights-UnifyoR@1, IoU=0.5: 20.33 -> 25.11Improved
use Segment query instead of Image query (ablation)expert model · QVHighlights-UnifyoR@1, IoU=0.5: 16.14 -> 20.33Improved
add Image-R query to textMoment-DETR (offline) · QVHighlights-Unify (offline)R@1, IoU=0.5: 53.94 -> 55.28Improved
add Segment-G query to textQD-DETR (offline) · QVHighlights-Unify (offline)R@1, IoU=0.5: 62.68 -> 64.03Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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