AI Paper Atlas

Research map / Video / Video-language understanding

VaTeX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research

Xin Wang, Jiawei Wu, Junkun Chen, Lei Li, Yuan-Fang Wang, William Yang Wang

ICCV 2019 · Video › Video-language understanding · approach: Video-language pretraining · 536 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 weight tying of embedding and softmax (ablation)Base captioning model · VaTeX Chinese captioning testBLEU-4: 24.4 -> 24.9Improved
add weight tying of embedding and softmax (ablation)Base captioning model · VaTeX English captioning testBLEU-4: 28.1 -> 28.1No clear effect
add weight tying of embedding and softmax (ablation)Base captioning model · VaTeX captioningparameters: 52.5M -> 39.7MImproved
share the video encoder across English and ChineseBase captioning model · VaTeX English captioning testCIDEr: 44.3 -> 45.1Improved
share the video encoder across English and ChineseBase captioning model · VaTeX captioningparameters: 39.7M -> 34.9MImproved
share encoder and decoder across languagesShared Enc · VaTeX captioningparameters: 39.7M -> 26.3MImproved
share encoder and decoder across languagesBase captioning model · VaTeX English captioning testBLEU-4: 28.1 -> 27.9No clear effect
add average video features (ablation)Base NMT · VaTeX English->Chinese translationBLEU-4: 26.85 -> 26.97No clear effect
add LSTM video features without attention (ablation)Base NMT · VaTeX English->Chinese translationBLEU-4: 26.85 -> 27.43Improved
add temporal attention over LSTM video features (ablation)LSTM VI w/o Attn · VaTeX English->Chinese translationBLEU-4: 27.43 -> 29.12Improved
add temporal attention over LSTM video features (ablation)LSTM VI w/o Attn · VaTeX Chinese->English translationBLEU-4: 24.76 -> 26.42Improved
add video context under 100% noun maskingNMT · VaTeX Chinese->English masked translationnoun recovery accuracy: 10.1 -> 39.7Improved

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