AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Stacked Cross Attention for Image-Text Matching

Kuang-Huei Lee, Xi Chen, Gang Hua, Houdong Hu, Xiaodong He

ECCV 2018 · Vision-language and multimodal › Human and embodied modeling · approach: Language-grounded scene text and structure · 1,435 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 stacked cross attention (SCAN t-i AVG vs Sum-Max t-i) (ablation)Sum-Max t-i · Flickr30KSentence R@1: 59.6 -> 61.8Improved
add stacked cross attention (SCAN i-t AVG vs Sum-Max i-t) (ablation)Sum-Max i-t · Flickr30KSentence R@1: 56.7 -> 67.9Improved
add stacked cross attention (SCAN i-t AVG vs Sum-Max i-t) (ablation)Sum-Max i-t · Flickr30KImage R@1: 36.8 -> 43.9Improved
infer latent region-word alignment (Sum-Max t-i vs VSE++ single embedding) (ablation)VSE++ · Flickr30KSentence R@1: 31.9 -> 59.6Improved
add hard negatives in triplet loss (ablation)SCAN i-t AVG · Flickr30KSentence R@1: 45.8 -> 67.9Improved
add hard negatives in triplet loss (ablation)SCAN i-t AVG · Flickr30KImage R@1: 33.9 -> 43.9Improved
normalize image embedding (ablation)SCAN i-t AVG · Flickr30KSentence R@1: 67.8 -> 67.9No clear effect
use AVG pooling instead of SUM (ablation)SCAN i-t · Flickr30KSentence R@1: 63.9 -> 67.9Improved
use AVG pooling instead of MAX (ablation)SCAN i-t · Flickr30KSentence R@1: 59.7 -> 67.9Improved
use SUM pooling instead of AVG (ablation)SCAN i-t · Flickr30KImage R@1: 43.9 -> 45.0Mixed
use bi-directional GRU instead of one-directional (ablation)SCAN i-t AVG · Flickr30KSentence R@1: 63.6 -> 67.9Improved
ensemble t-i AVG and i-t LSE modelsSCAN single models · Flickr30KImage R@1: 45.8 -> 48.6Improved

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