AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DAS: Densely-Anchored Sampling for Deep Metric Learning

Lizhao Liu, Shangxin Huang, Zhuangwei Zhuang, Ran Yang, Mingkui Tan, Yaowei Wang

ECCV 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Correspondence and causal representation learning · 22 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 DAS to ResNet50-512 margin-style MSMS · CUBRecall@1 (%): 66.46 -> 69.19Improved
add DAS to ResNet50-512 MSMS · CARSRecall@1 (%): 83.99 -> 87.84Improved
add DAS to triplet loss with distance weighted samplingTriplet [D] · CARSRecall@1 (%): 78.86 -> 82.63Improved
add DAS to contrastive loss with distance samplingContrastive [D] · CARSRecall@1 (%): 76.03 -> 80.74Improved
add DAS to whole-batch loss GenLiftedGenLifted · CARSRecall@1 (%): 72.45 -> 73.55Improved
add DAS to triplet loss with random samplingTriplet random sampling · CARSRecall@1 (%): 74.21 -> 76.79Improved
add DAS to contrastive loss with random samplingContrastive random sampling · CARSRecall@1 (%): 42.44 -> 50.79Improved
add DFS module (ablation)Triplet [D] + DAS · CARSRecall@1 (%): 78.86 -> 81.28Improved
add MTS module (ablation)Triplet [D] + DAS · CARSNMI: 65.85 -> 67.81Improved
add MTS on top of DFS (ablation)Triplet [D] + DAS · CARSRecall@1 (%): 81.28 -> 82.63Improved
increase K in DFS beyond 4 (ablation)DFS with margin loss · SOPRecall@1 (%): 79.14 -> 78.09Hurt
increase bank capacity Z in MTS (ablation)MTS with margin loss · CARSRecall@1 (%): 82.35 -> 82.75Improved

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