AI Paper Atlas

Research map / Video / Video-language understanding

MultiVENT: Multilingual Videos of Events and Aligned Natural Text

Kate Sanders, David Etter, Reno Kriz, Benjamin Van Durme

NeurIPS 2023 · Video › Video-language understanding · approach: Video-language pretraining · 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 multilingual pretraining (XLM-R text encoder, LAION-5B) to pooled CLIPpooled CLIP · MultiVENT non-EnglishRecall@1: 6.2 -> 32.6Improved
add multilingual pretraining to pooled CLIPpooled CLIP · MultiVENT non-EnglishRecall@10: 22.4 -> 79.5Improved
add multilingual pretraining to pooled CLIPpooled CLIP · MultiVENT non-EnglishPrecision@10 (event description): 58.2 -> 61.5Improved
add multilingual pretraining to pooled CLIPpooled CLIP · MultiVENT EnglishRecall@1: 55.9 -> 55.9No clear effect
add multilingual pretraining to pooled CLIPpooled CLIP · MultiVENT EnglishRecall@10: 91.3 -> 92.3No clear effect
evaluate on non-English instead of English videosInternVideo · MultiVENTRecall@1: 53.8 -> 5.7Hurt
evaluate on non-English instead of English videosMultiCLIP · MultiVENTRecall@1: 55.9 -> 32.6Hurt
evaluate on non-English instead of English videosCLIP2Video · MultiVENTRecall@1: 41.3 -> 2.4Hurt
evaluate on non-English videos with English event queriesMultiCLIP · MultiVENTPrecision@10: 80.6 -> 61.5Hurt
use news broadcast instead of raw footage videosannotation · MultiVENT disaster videosentities per video: 3.63 -> 13.78Improved
use news broadcast instead of edited footageannotation · MultiVENT disaster videoson-screen text entities per video: 2.36 -> 7.87Improved
use MultiCLIP instead of Clip2Video on MSR-VTTClip2Video · MSR-VTTRecall@1: 29.8 -> 38.4Improved

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