AI Paper Atlas

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

Counterfactual Samples Synthesizing and Training for Robust Visual Question Answering

Long Chen, Yuhang Zheng, Yulei Niu, Hanwang Zhang, Jun Xiao

TPAMI 2023 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 57 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 CSST-L contrastive trainingUpDn · VQA-CP v2 testaccuracy (All): 39.68 -> 56.55Improved
add CSST-G contrastive trainingLMH · VQA-CP v2 testaccuracy (All): 52.45 -> 61.66Improved
use CSST-L instead of CSST-G (ablation)LMH · VQA-CP v2 testaccuracy (All): 61.66 -> 55.38Hurt
add CSST-G contrastive trainingRUBi · VQA-CP v2 testaccuracy (All): 45.23 -> 47.06Improved
add CSST contrastive loss on top of CSS+ (ablation)LMH · VQA-CP v2 testaccuracy (All): 59.54 -> 61.66Improved
add CSST contrastive loss on top of CSS+ (ablation)LMH (LXMERT) · VQA-CP v2 testaccuracy (All): 63.63 -> 65.71Improved
add CSS+ synthesized samples (ablation)LMH · VQA-CP v2 testaccuracy (All): 52.45 -> 59.54Improved
add improved CSS+ in place of original CSS (ablation)LMH · VQA-CP v2 testaccuracy (All): 58.95 -> 59.54Improved
add SAR rerankingLMH-CSST · VQA-CP v1 testaccuracy (All): 63.16 -> 69.57Improved
add CSST-style CSS-negative selection instead of random-only negatives (ablation)LMH CR · VQA-CP v2 testaccuracy (All): 60.74 -> 61.66Improved
add CSST selection in place of prior CL strategy (ablation)LMH-CSS+ · VQA-CP v2 testaccuracy (All): 59.78 -> 61.66Improved
add CSS+ synthesized samples (ablation)RUBi · GQA-OOD testdevDelta (imbalance): 9.2 -> 14.5Mixed

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