AI Paper Atlas

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

A-OKVQA: A Benchmark for Visual Question Answering Using World Knowledge

Dustin Schwenk, Apoorv Khandelwal, Christopher Clark, Kenneth Marino, Roozbeh Mottaghi

ECCV 2022 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 414 citations (OpenAlex)

Read the paper

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 ground-truth rationales as input (ablation)GPV-2 · A-OKVQA multiple choice (test)accuracy: 53.7 -> 67.2Improved
add ground-truth rationales as input (ablation)GPV-2 · A-OKVQA direct answer (test)accuracy: 40.7 -> 51.7Improved
mask the answer inside ground-truth rationales (vs unmasked) (ablation)GPV-2 + GT Ratl. · A-OKVQA multiple choice (test)accuracy: 67.2 -> 58.3Hurt
mask answer in rationales vs no rationales (ablation)GPV-2 · A-OKVQA multiple choice (test)accuracy: 53.7 -> 58.3Improved
replace GT captions with GT rationales in the GPT-3 prompt (ablation)GPT-3 · A-OKVQA multiple choice (val)accuracy: 45.40 -> 56.74Improved
replace generated captions with generated rationales in the GPT-3 prompt (ablation)ClipCap -> GPT-3 · A-OKVQA multiple choice (test)accuracy: 43.61 -> 43.84No clear effect
add image input to question-only CLIP classifier (ablation)CLIP classifier · A-OKVQA direct answer (test)accuracy: 10.24 -> 14.27Improved
fine-tune ViLBERT on A-OKVQA after VQA training vs on A-OKVQA alone, direct answer (ablation)ViLBERT · A-OKVQA direct answer (test)accuracy: 25.9 -> 12.0Hurt
train CLIP contrastive MLP instead of zero-shot, image onlyCLIP image-only · A-OKVQA multiple choice (test)accuracy: 53.94 -> 50.09Hurt
add GPV-2 to the majority-vote ensemblemajority vote · A-OKVQA direct answer (test)accuracy: 35.60 -> 40.7Hurt
majority vote of five weaker models vs best single of themmajority vote · A-OKVQA direct answer (test)accuracy: 27.1 -> 29.5Improved
oracle selection across six models vs best single modelmodel ensemble · A-OKVQA direct answer (test)accuracy: 40.7 -> 56.87Improved

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