AI Paper Atlas

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

RUBi: Reducing Unimodal Biases for Visual Question Answering

Remi Cadene, Corentin Dancette, Hedi Ben younes, Matthieu Cord, Devi Parikh

NeurIPS 2019 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 66 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
Apply RUBi to the baseline architectureBaseline architecture · VQA-CP v2 testOverall accuracy: 38.46 $\pm$ 0.07 -> 47.11 $\pm$ 0.51Improved
Apply RUBi to SANSAN · VQA-CP v2 testOverall accuracy: 24.96 -> 37.63Improved
Apply RUBi to UpDnUpDn · VQA-CP v2 testOverall accuracy: 39.74 -> 44.23Improved
Balance the answer distribution within each question typeBaseline architecture · VQA-CP v2 testOverall accuracy: 38.46 -> 42.11Improved
Apply RUBi to the baseline architectureBaseline architecture · VQA v2 valOverall accuracy: 63.10 -> 61.16Hurt
Apply RUBi to the baseline architectureBaseline architecture · VQA v2 test-devOverall accuracy: 64.75 -> 63.18Hurt
Replace sigmoid masking with ReLU (ablation)RUBi · VQA-CP v2 testAccuracy: -7.09 accuracy pointHurt
Use element-wise sum instead of element-wise product for fusion (ablation)RUBi · VQA-CP v2 testAccuracyHurt
Add the question-only loss L_QO (ablation)Baseline architecture + RUBi · VQA-CP v2 testOverall accuracy: 46.11 -> 47.11Improved
Add the question-only loss L_QO (ablation)SAN + RUBi · VQA-CP v2 testOverall accuracy: 36.96 -> 37.63Improved
Add the question-only loss L_QO (ablation)UpDn + RUBi · VQA-CP v2 testOverall accuracy: 39.47 -> 44.23Improved
Apply RUBi to the baseline architectureBaseline architecture · VQA-CP v1Overall accuracy: 37.13 -> 46.93Improved

Later papers that compared with this method

7 later accepted papers report RUBi as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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