AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Vx2Text: End-to-End Learning of Video-Based Text Generation From Multimodal Inputs

Xudong Lin, Gedas Bertasius, Jue Wang, Shih-Fu Chang, Devi Parikh, Lorenzo Torresani

CVPR 2021 · Vision-language and multimodal › Cross-modal learning · approach: Audio-visual speech separation · 54 citations (OpenAlex)

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 Vx2Text instead of SimpleAVSD captioning without caption input · AVSD testCIDEr: 0.905 -> 1.357Improved
use Vx2Text instead of MTN-TMTAVSD captioning with caption input · AVSD testCIDEr: 1.357 -> 1.605Improved
use Vx2Text instead of BERT QATVQA · TVQA validationTop-1 accuracy: 72.4 -> 74.9Improved
use Vx2Text instead of HERO with multimodal pretrainingTVQA · TVQA testTop-1 accuracy: 73.6 -> 75.0Improved
use Vx2Text instead of MMTTVC video captioning · TVC validationCIDEr: 0.444 -> 0.482Improved
add video input (ablation)Vx2Text · AVSD validationperformanceImproved
add each available modality (ablation)Vx2Text · AVSD and TVQA validationperformanceImproved
add previous question-answer history (ablation)Vx2Text · AVSD validationperformanceImproved
use differentiable tokenization instead of frozen tokenization (ablation)Vx2Text · AVSD and TVQA validationperformanceImproved
use frozen tokenization instead of multimodal feature embedding (ablation)Vx2Text · AVSD and TVQA validationperformanceImproved
use generative decoding instead of a discriminative classification head (ablation)Vx2Text · TVQA with 10% of training dataTop-1 accuracy: 34.2% -> 64.1%Improved
add cycle-consistency training (ablation)Generative Vx2Text · TVQA with 10% of training dataTop-1 accuracy: 64.1% -> 66.1%Improved

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