AI Paper Atlas

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

Deep Spatial-Semantic Attention for Fine-Grained Sketch-Based Image Retrieval

Jifei Song, Qian Yu, Yi-Zhe Song, Tao Xiang, Timothy M. Hospedales

ICCV 2017 · Vision-language and multimodal › Human and embodied modeling · approach: Composed and text-based image retrieval · 268 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 coarse-fine fusion (CFF) (ablation)Triplet SN base · QMUL-Shoeacc@1: 52.17% -> 58.26%Improved
add coarse-fine fusion (CFF) (ablation)Triplet SN base · QMUL-Chairacc@10: 98.96% -> 95.88%Hurt
add HOLEF loss (ablation)Triplet SN base · QMUL-Shoeacc@1: 52.17% -> 56.52%Improved
add HOLEF loss (ablation)Triplet SN base · Our Handbagacc@1: 39.88% -> 40.48%No clear effect
add CFF and HOLEF together (ablation)Triplet SN base · QMUL-Chairacc@1: 72.16% -> 81.44%Improved
add attention module (ablation)full model · QMUL-Chairacc@1: 77.32% -> 81.44%Improved
add attention module (ablation)full model · QMUL-Shoeacc@1: 58.26% -> 61.74%Improved
add attention module (ablation)Base + HOLEF · Our Handbagacc@1: 40.48% -> 40.48%No clear effect
add shortcut connection in attention module (ablation)full model · QMUL-Shoeacc@1: 27.83% -> 61.74%Improved
add shortcut connection in attention module (ablation)full model · QMUL-Chairacc@1: 49.48% -> 81.44%Improved
use HOLEF instead of Euclidean triplet loss (ablation)full model · QMUL-Shoeacc@1: 58.26% -> 61.74%Improved
use HOLEF instead of Mahalanobis triplet loss (ablation)full model · Our Handbagacc@1: 44.64% -> 49.40%Improved

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