AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Switch-BERT: Learning to Model Multimodal Interactions by Switching Attention and Input

Qingpei Guo, Kaisheng Yao, Wei Chu

ECCV 2022 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 7 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 Switch-Input and Switch-Attention blocks to plain single-stream BERT (ablation)BERT encoder · VQAv2 test-devVQA-score: 67.8 -> 70.7Improved
add Switch-Input and Switch-Attention blocks to plain single-stream BERT (ablation)BERT encoder · RefCOCO+accuracy: 69.5 -> 72.8Improved
add Switch-Attention Block alone to plain BERT (ablation)BERT encoder · VQAv2 test-devVQA-score: 67.8 -> 68.9Improved
add Switch-Input Block to SAB-only model (ablation)Switch-BERT · VQAv2 test-devVQA-score: 68.9 -> 70.7Improved
add Switch-Input Block to SAB-only model (ablation)Switch-BERT · RefCOCO+accuracy: 72.4 -> 72.8Improved
add Switch-Attention Block to SIB-only model (ablation)Switch-BERT · Flickr30k text retrievalRecall@1: 77.7 -> 78.2Improved
add Joint-Attention mode to Cross-Self and Self-Cross modes (pretrained) (ablation)Switch-Attention Block · RefCOCO+accuracy: 71.5 -> 72.8Improved
add Joint-Attention mode to Cross-Self and Self-Cross modes (no pretraining) (ablation)Switch-Attention Block · RefCOCO+accuracy: 68.3 -> 68.9Improved
add Cross-Self and Self-Cross modes to Joint-Attention (pretrained) (ablation)Switch-Attention Block · RefCOCO+accuracy: 71.2 -> 72.8Improved
add pretraining on Conceptual Captions (ablation)Switch-BERT · Flickr30k image retrievalRecall@1: 38.2 -> 62.2Improved
add pretraining on Conceptual Captions (ablation)Switch-BERT · VQAv2 test-devVQA-score: 66.7 -> 70.7Improved
increase depth from 6 to 12 layers (ablation)Switch-BERT · VQAv2 test-devVQA-score: 65.4 -> 70.7Improved

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