AI Paper Atlas

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

FVQA: Fact-Based Visual Question Answering

Peng Wang, Qi Wu, Chunhua Shen, Anthony Dick, Anton van den Hengel

TPAMI 2017 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 512 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
use top-3 instead of top-1 predicted question-query mappings (ablation)FVQA reasoning model · FVQA overall Top-1accuracy %: 52.56 -> 56.91Improved
use ground-truth question-query mapping instead of predicted top-3 (ablation)FVQA reasoning model · FVQA overall Top-1accuracy %: 56.91 -> 63.63Improved
use explicit KB reasoning instead of LSTM-Question+ImageLSTM baseline · FVQA overall Top-1accuracy %: 22.97 -> 56.91Improved
ensemble LSTM model with knowledge-based modelFVQA reasoning model · FVQA overall Top-1accuracy %: 56.91 -> 58.76Improved
add image input to question-only LSTMLSTM-Question · FVQA overall Top-1accuracy %: 10.45 -> 22.97Improved
add Pre-VQA extra training data to hierarchical modelHie-Question+Image · FVQA overall Top-1accuracy %: 33.70 -> 43.14Improved
add Pre-VQA extra training data to LSTM modelLSTM-Question+Image · FVQA overall Top-1accuracy %: 22.97 -> 24.98Improved
add question to SVM image featuresSVM-Image · FVQA overall Top-1accuracy %: 18.41 -> 18.89No clear effect
use top-3 mapping on supporting-fact prediction (ablation)FVQA reasoning model · FVQA supporting-fact Top-1accuracy %: 38.76 -> 41.12Improved
compare WebChild questions with other KB sources for mappingQQmaping · FVQATop-1 mapping accuracy %: 83.15 -> 64.94Improved
proposed model vs Hie+Pre-VQA on WebChild questionsFVQA reasoning model · FVQA WebChild Top-1accuracy %: 48.74 -> 52.85Hurt

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