AI Paper Atlas

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

Robust Visual Reasoning via Language Guided Neural Module Networks

Arjun Reddy Akula, Varun Jampani, Soravit Changpinyo, Song-Chun Zhu

NeurIPS 2021 · Vision-language and multimodal › Visual question answering · approach: Vision-language model VQA evaluation · 5 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
replace LG-Conv with standard convolution (ablation)Ours (with CoSAtt) · CLOSUREaccuracy: 88.0 -> 62.1Hurt
replace LG-Conv with standard convolution (ablation)Ours (with CoSAtt) · S-Ref Testaccuracy: 63.3 -> 51.6Hurt
replace LG-Conv with standard convolution (ablation)Ours (with CoSAtt) · F-Ref Testaccuracy: 84.3 -> 77.8Hurt
add image conditioning to the CoSAtt guidance (ablation)Ours (with CoSAtt) · CLOSUREaccuracy: 80.2 -> 88.0Improved
add query conditioning to the CoSAtt guidance (ablation)Ours (with CoSAtt) · CLOSUREaccuracy: 78.8 -> 88.0Improved
add module argument to the CoSAtt guidance (ablation)Ours (with CoSAtt) · CLOSUREaccuracy: 82.1 -> 88.0Improved
add module argument to the CoSAtt guidance (ablation)Ours (with CoSAtt) · S-Ref Testaccuracy: 61.7 -> 63.3Improved
add module argument to the CoSAtt guidance (ablation)Ours (with CoSAtt) · F-Ref Testaccuracy: 80.9 -> 84.3Improved
use CoSAtt encoder instead of BiSAtt encoderOurs · CLOSUREaccuracy: 86.1 -> 88.0Improved
use CoSAtt encoder instead of BiSAtt encoderOurs · S-Ref Testaccuracy: 59.7 -> 63.3Improved
use CoSAtt encoder instead of BiSAtt encoderOurs · CLEVR-Testaccuracy: 99.2 -> 99.2No clear effect
evaluate on C3-Ref+ novel compositions and contrast sets versus S-RefOurs · C3-Ref+accuracyHurt

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