HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Image Retrieval On Sop

평가 지표

R@1

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	R@1	Paper Title	Repository
Cross-Batch Memory	80.6	Cross-Batch Memory for Embedding Learning
Unicom+ViT-L@336px	91.2	Unicom: Universal and Compact Representation Learning for Image Retrieval
ROADMAP (DeiT-B)	86.0	Robust and Decomposable Average Precision for Image Retrieval
MS512	78.2	Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning
ABE-8	76.3	Attention-based Ensemble for Deep Metric Learning	-
ROADMAP (ResNet-50)	83.1	Robust and Decomposable Average Precision for Image Retrieval
EPSHN512	78.3	Improved Embeddings with Easy Positive Triplet Mining
Smooth-AP	80.1	Smooth-AP: Smoothing the Path Towards Large-Scale Image Retrieval
PNP Loss	81.1	Rethinking the Optimization of Average Precision: Only Penalizing Negative Instances before Positive Ones is Enough
ProxyNCA++	81.4	ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis
HDC	69.5	Hard-Aware Deeply Cascaded Embedding
CGD (SG/GS)	84.2	Combination of Multiple Global Descriptors for Image Retrieval
NormSoftmax2048 (ResNet-50)	79.5	Classification is a Strong Baseline for Deep Metric Learning
A-BIER	74.2	Deep Metric Learning with BIER: Boosting Independent Embeddings Robustly

0 of 14 row(s) selected.