AI Paper Atlas

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

Structured Attentions for Visual Question Answering

Chen Zhu, Yanpeng Zhao, Shuaiyi Huang, Kewei Tu, Yi Ma

ICCV 2017 · Vision-language and multimodal › Visual question answering · approach: Attention and prior-robust VQA · 118 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 MF structured attention to sigmoid attention (ablation)SIG · CLEVR validation (res5c)overall accuracy: 70.52 -> 73.14Improved
add LBP structured attention to sigmoid attention (ablation)SIG · CLEVR validation (res5c)overall accuracy: 70.52 -> 72.30Improved
add MF structured attention (ablation)SIG · CLEVR validation (res4b22)overall accuracy: 75.32 -> 76.65Improved
combine LBP with SIG attention (ablation)LBP · CLEVR validation (res4b22)overall accuracy: 76.21 -> 77.97Improved
combine MF with SIG attention (ablation)MF · CLEVR validation (res5c)overall accuracy: 73.14 -> 73.19No clear effect
use MF-SIG instead of SANVQA model · CLEVR testoverall accuracy: 68.5 -> 77.57Improved
use res4b22 instead of res5c featuresSIG · CLEVR validationoverall accuracy: 70.52 -> 75.32Improved
increase inference steps T from 1 to 3 (ablation)MF-SIG · VQA test-devoverall accuracy: 65.90 -> 66.00No clear effect
increase inference steps T from 1 to 4 (ablation)MF-SIG · VQA test-devoverall accuracy: 65.90 -> 65.81No clear effect
increase inference steps T from 1 to 3 (ablation)LBP-SIG · VQA test-devoverall accuracy: 65.93 -> 65.81Hurt
add Visual Genome external dataMF-SIG · VQA test-devoverall accuracy: 66.00 -> 67.19Improved
use MF-SIG instead of MLB under VG dataVQA model · VQA test-devoverall accuracy: 65.84 -> 67.19Improved

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