AI Paper Atlas

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

Beyond Question-Based Biases: Assessing Multimodal Shortcut Learning in Visual Question Answering

Corentin Dancette, Rémi Cadène, Damien Teney, Matthieu Cord

ICCV 2021 · Vision-language and multimodal › Visual question answering · approach: Universal VQA models · 92 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 LfF debiasingUpDown · VQA v2 Counterexamplesaccuracy: 33.91 -> 34.27Improved
add RandImg debiasingUpDown · VQA v2 Counterexamplesaccuracy: 33.91 -> 34.41Improved
add LMH debiasingUpDown · VQA v2 Counterexamples vs Easyaccuracy: 33.91 -> 34.26 (Counterexamples), 76.69 -> 73.12 (Easy)Mixed
add LMH+CSS debiasingUpDown · VQA v2 Counterexamples vs Easyaccuracy: 33.91 -> 34.36 (Counterexamples), 76.69 -> 62.08 (Easy)Mixed
add RUBi debiasingUpDown · VQA v2 Counterexamplesaccuracy: 33.91 -> 32.25Hurt
add LMH+RMFE debiasingUpDown · VQA v2 overallaccuracy: 63.52 -> 60.96Hurt
add ESR debiasingUpDown · VQA v2 overallaccuracy: 63.52 -> 62.96Hurt
use BLOCK bilinear fusion instead of UpDownUpDown · VQA v2 Counterexamplesaccuracy: 33.91 -> 32.91Hurt
use BLOCK bilinear fusion instead of UpDownUpDown · VQA v2 Easyaccuracy: 76.69 -> 77.65Improved
add VilBERT pretraining on Conceptual CaptionsUpDown-style models · VQA v2 Counterexamplesaccuracy: 33.91 -> 39.24Improved
move from overall to Counterexamples subsetVQA models SAN, UpDown, BLOCK, VilBERT · VQA v2accuracy: about 29 points lostHurt
apply bias-reduction methods trained for VQA-CP to natural VQA v2 shortcutsUpDown · VQA-CP v2 vs VQA-CEaccuracy: +5 to +20 on VQA-CP, gains of +0.36 and +0.50 at best on CounterexamplesMixed

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