AI Paper Atlas

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

TRAR: Routing the Attention Spans in Transformer for Visual Question Answering

Yiyi Zhou, Tianhe Ren, Chaoyang Zhu, Xiaoshuai Sun, Jianzhuang Liu, Xinghao Ding, Mingliang Xu, Rongrong Ji

ICCV 2021 · Vision-language and multimodal › Visual question answering · approach: Vision-language model VQA evaluation · 133 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 TRAR attention-span routing (soft) to the multi-modal TransformerMCAN · VQA2.0 valaccuracy (All): 67.3 -> 67.7Improved
add TRAR attention-span routing (soft) to the one-stage REC TransformerDarkNet + Transformer · RefCOCO valIoU accuracy: 75.8 -> 77.6Improved
use mask-based routing (TRAR) instead of conventional multi-module routingMCAN · VQA2.0 valTransformer parameters: 67M -> 45MImproved
use conventional (separate-module) routing instead of the plain TransformerDarkNet + Transformer · RefCOCO valIoU accuracy: 75.8 -> 75.1Hurt
add a 1st-order local attention mask to the routing space (ablation)TRAR-S · VQA2.0 valaccuracy (All): 67.3 -> 67.6Improved
add a 1st-order local attention mask to the routing space (ablation)TRAR-S · RefCOCO valIoU accuracy: 75.8 -> 76.2Improved
add a 3rd-order mask (routing space [0,1,2] -> [0,1,2,3]) (ablation)TRAR-S · VQA2.0 valaccuracy (All): 67.7 -> 67.6No clear effect
add a 3rd-order mask (routing space [0,1,2] -> [0,1,2,3]) (ablation)TRAR-S · RefCOCO valIoU accuracy: 76.7 -> 77.6Improved
add a 3rd-order mask (routing space [0,1,2] -> [0,1,2,3]) (ablation)TRAR-H · RefCOCO valIoU accuracy: 76.9 -> 76.1Hurt
add a 4th-order mask (routing space [0,1,2,3] -> [0,1,2,3,4]) (ablation)TRAR-S · RefCOCO valIoU accuracy: 77.6 -> 77.5No clear effect
use 16x16 grid features instead of the 8x8 resolutionTRAR-S · VQA2.0 test-devaccuracy (All): 72.00 -> 72.62Improved
add TRAR (soft) to the Transformer on CLEVRMCAN · CLEVRoverall accuracy: 98.54 -> 99.00Improved

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