AI Paper Atlas

Research map / Video / Video-language understanding

QSVideo: Query-Conditioned Semantic Temporal Retrieval for Video Understanding

Wei Ao, Lan Wang, Vishnu Boddeti

ECCV 2026 · Video › Video-language understanding · approach: Long video understanding · 0 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 QSVideo retrieval to MiniCPM-o 2.6 (16 frames)MiniCPM-o 2.6 · LVBenchoverall accuracy (%): 38.9 -> 45.8Improved
add QSVideo retrieval to MiniCPM-o 2.6 (32 frames)MiniCPM-o 2.6 · LVBenchoverall accuracy (%): 42.0 -> 49.1Improved
add QSVideo retrieval to MiniCPM-o 2.6 (8 frames)MiniCPM-o 2.6 · StreamingBench real-timeoverall accuracy (%): 64.44 -> 77.68Improved
add QSVideo retrieval to MiniCPM-o 2.6 (16 frames)MiniCPM-o 2.6 · StreamingBench real-timeoverall accuracy (%): 65.12 -> 79.36Improved
plug QSVideo into MiniCPM-V 2.6 at 8 framesMiniCPM-V 2.6 · StreamingBench real-timeoverall accuracy (%): 66.36 -> 70.96Improved
plug QSVideo into LLaVA-OV, gain shrinks with more framesLLaVA-OV · StreamingBench real-timeoverall accuracy (%): 72.12 -> 74.48 (8 frames); 73.56 -> 74.40 (32 frames)Improved
replace holistic single-score ranker with structured object/action/location QSRanker (ablation)QSRanker · LVBench ranking Recall@100Recall@100 (%): 52.78 -> 58.20Improved
replace holistic ranker with QSRanker at top-1 retrieval (ablation)QSRanker · LVBench ranking Recall@1Recall@1 (%): 17.38 -> 17.51No clear effect
increase lambda (relevance weight) from 0.1 to 0.9 at 16 frames (ablation)QSRetrieval · LVBench (16 frames)overall accuracy (%): 46.1 (0.1) vs 45.8 (0.9)No clear effect
increase lambda from 0.1 to 0.9 at 32 frames (ablation)QSRetrieval · LVBench (32 frames)overall accuracy (%): 48.2 -> 49.1Improved
use batch inference and data parallelism for semantic ranker (ablation)QSRanker · latency breakdownlatency (s): 608.40s -> 61.32sImproved
use QSVideo with 8B VLM instead of SEAL 34BQSVideo · LVBench end-to-end latencylatency (s): 1702.24 -> 88.73Improved

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