AI Paper Atlas

Research map / Video / Video-language understanding

OmniCVR: A Benchmark for Omni-Composed Video Retrieval with Vision, Audio, and Text

Junyang Ji, Shengjun Zhang, Da Li, Yuxiao Luo, Yan Wang, Di Xu, Biao Yang, Wei Yuan, Fan Yang, Zhihai He, Wenming Yang

ICLR 2026 · Video › Video-language understanding · approach: Text-video retrieval

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 audio-to-text descriptions from Qwen2-AudioVLM2Vec · OmniCVR audio-centricR@1: 12.4 -> 77.2Improved
add audio-to-text descriptions from Qwen2-AudioVLM2Vec · OmniCVR (overall)R@1: 38.44 -> 66.98Improved
add audio-to-text descriptions from Qwen2-AudioVLM2Vec · OmniCVR (vision-centric)R@1: 55.04 -> 69.33Improved
add audio-to-text descriptions, human speech targetsVLM2Vec · OmniCVR audio-centricR@1: 11.36 -> 96.59Improved
add audio-to-text descriptions, music targetsVLM2Vec · OmniCVR audio-centricR@1: 16.07 -> 86.43Improved
add audio-to-text descriptions, sound targetsVLM2Vec · OmniCVR audio-centricR@1: 10.75 -> 60.31Improved
add source video frames as input (ablation)AudioVLM2Vec · OmniCVR audio-centricR@1: 28.10 -> 77.20Improved
replace native audio tokens with Audio-as-Text captions (ablation)OmniEmbed-v0.1-multivent · OmniCVR audio-centricR@1: 13.6 -> 32.7Improved
replace native audio tokens with Audio-as-Text captions, speech (ablation)OmniEmbed-v0.1-multivent · OmniCVR audio-centricR@1: 11.36 -> 50.38Improved
replace native audio tokens with Audio-as-Text captions, music (ablation)OmniEmbed-v0.1-multivent · OmniCVR audio-centricR@1: 16.07 -> 30.00Improved
add audio-to-text descriptions zero-shotVLM2Vec · MSR-VTT zero-shot T2VR@1: 36.10 -> 37.90Improved
add audio captioning step (latency cost)VLM2Vec · OmniCVR audio-centricinference latency (s): 1.72 -> 4.77Hurt

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