AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Co-Domain Embedding Using Deep Quadruplet Networks for Unseen Traffic Sign Recognition

Junsik Kim, Seokju Lee, Tae-Hyun Oh, In So Kweon

AAAI 2018 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Fine-grained and cross-domain few-shot baselines · 23 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
Use template-anchored quadruplet training with HingeM-5 instead of Triplet-DATraffic-sign one-shot recognition · GTSRB-sub unseen classesTop 1-NN accuracy (%): 47.6 -> 47.8No clear effect
Use template-anchored quadruplet training with HingeM-5 instead of Triplet-DATraffic-sign one-shot recognition · TT100K-c unseen classesTop 1-NN accuracy (%): 67.1 -> 67.3No clear effect
Use template-anchored quadruplet training with HingeM-5 instead of IdsiaNetTraffic-sign one-shot recognition · GTSRB unseen classesTop 1-NN accuracy (%): 40.2 -> 45.2Improved
Train quadruplet embeddings with HingeM instead of IdsiaNet featuresCross-dataset traffic-sign transfer · TT100K-c sample average (trained on GTSRB)Top1 NN accuracy (%): 36.5 -> 42.3Improved
Use 3 pairwise loss terms instead of HingeM-5 (ablation)Quadruplet network · GTSRB-all validation setTop1 NN accuracy, Avg. (%): 69.1 -> 67.3Hurt
Add real-real pairwise loss term to HingeM-5 (ablation)Quadruplet network · GTSRB-all validation setTop1 NN accuracy, Unseen (%): 41.6 -> 39.2Hurt
Add real-real pairwise loss term to HingeM-5 (ablation)Quadruplet network · GTSRB-all validation setTop1 NN accuracy, Seen (%): 95.3 -> 97.3Improved
Set embedding dimension to 50 instead of 100 (ablation)Quadruplet network with HingeM-5 · GTSRB-all validation setTop1 NN accuracy, Avg. (%): 69.1 -> 67.1Hurt
Set embedding dimension to 150 instead of 100 (ablation)Quadruplet network with HingeM-5 · GTSRB-all validation setTop1 NN accuracy, Avg. (%): 69.1 -> 68.9Hurt
Replace HingeM pairwise losses with contrastive-style losses (ablation)Quadruplet network · GTSRB-all unseen-class representation, 200 SVM instances/classSVM classification error (%): 3.77 ($\pm$0.29) -> 5.30 ($\pm$0.14)Hurt
Replace HingeM pairwise losses with contrastive-style losses (ablation)Quadruplet network · GTSRB-sub unseen-class representation, 200 SVM instances/classSVM classification error (%): 13.81 ($\pm$0.31) -> 12.12 ($\pm$0.33)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