AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Towards Video Text Visual Question Answering: Benchmark and Baseline

Minyi Zhao, Bingjia Li, Jie Wang, Wanqing Li, Wenjing Zhou, Lan Zhang, Shijie Xuyang, Zhihang Yu, Xinkun Yu, Guangze Li, Aobotao Dai, Shuigeng Zhou

NeurIPS 2022 · Vision-language and multimodal › Visual question answering · approach: Multimodal question answering datasets · 1 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 OCR features (text, visual, temporal) on top of video features (ablation)T5-ViteVQA · M4-ViteVQA Task1Split1 valaccuracy: 11.51 -> 23.17Improved
add OCR visual features (ablation)T5-ViteVQA · M4-ViteVQA Task1Split1 valaccuracy: 20.79 -> 23.17Improved
add OCR temporal features (start/end temporal ids) (ablation)T5-ViteVQA · M4-ViteVQA Task1Split1 valaccuracy: 19.78 -> 23.17Improved
add video features Xv (ablation)T5-ViteVQA · M4-ViteVQA Task1Split1 valaccuracy: 22.36 -> 23.17Improved
use T5-ViteVQA instead of M4C (temporal reasoning)T5-ViteVQA vs M4C · M4-ViteVQA Task1Split1 valaccuracy: 17.91 -> 23.17Improved
use T5-ViteVQA instead of M4C on easy questionsT5-ViteVQA vs M4C · M4-ViteVQA Task1Split1 valaccuracy: 19.30 -> 25.09Improved
use T5-ViteVQA instead of M4C on hard questionsT5-ViteVQA vs M4C · M4-ViteVQA Task1Split1 valaccuracy: 9.02 -> 14.26Improved
use T5-ViteVQA instead of M4C on text questionsT5-ViteVQA vs M4C · M4-ViteVQA Task1Split1 valaccuracy: 17.26 -> 23.08Improved
use T5-ViteVQA instead of M4C on vision questionsT5-ViteVQA vs M4C · M4-ViteVQA Task1Split1 valaccuracy: 18.36 -> 24.21Improved
use T5-ViteVQA instead of M4C on harder split (cross-video-source)T5-ViteVQA vs M4C · M4-ViteVQA Task1Split2 testaccuracy: 11.36 -> 16.68Improved
use T5-ViteVQA instead of M4C in domain adaptation Task2 without adaptationT5-ViteVQA vs M4C · M4-ViteVQA Task2 testaccuracy: 7.52 -> 9.29Improved
use T5-ViteVQA instead of the VideoQA model JustAskT5-ViteVQA vs JuskAsk · M4-ViteVQA Task1Split1 valaccuracy: 10.81 -> 23.17Improved

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