Image Retrieval on Revisited Oxford (ROxf) (Medium)
91.2mAPSuperGlobal
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SuperGlobalBackbone=ResNet-101, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=8002023.08 | 91.2 | — | |
| SuperGlobalBackbone=ResNet-101, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=16002023.08 | 91.2 | — | |
| SuperGlobalBackbone=ResNet-101, Retrieval Category=SuperGlobal retrieval and reranking, Reranking Setting=Rerank top 4002023.08 | 90.9 | — | |
| SuperGlobalBackbone=ResNet-101, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=4002023.08 | 90.9 | — | |
| SuperGlobalBackbone=ResNet-50, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=8002023.08 | 88.9 | — | |
| SuperGlobalBackbone=ResNet-50, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=16002023.08 | 88.9 | — | |
| SuperGlobalBackbone=ResNet-50, Retrieval Category=SuperGlobal retrieval and reranking, Reranking Setting=Rerank top 4002023.08 | 88.8 | — | |
| SuperGlobalBackbone=ResNet-50, Strategy=SuperGlobal feature retrieval and reranking, Reranking Depth=4002023.08 | 88.8 | — | |
| CVNetBackbone=ResNet-50, Retrieval Category=Global feature retrieval + Local feature reranking, Reranking Setting=Rerank top 4002023.08 | 87.9 | — | |
| CVNetBackbone=ResNet-50, Strategy=Global feature retrieval + Local feature reranking, Reranking Depth=4002023.08 | 87.9 | — | |
| CVNetBackbone=ResNet-101, Retrieval Category=Global feature retrieval + Local feature reranking, Reranking Setting=Rerank top 4002023.08 | 87.2 | — | |
| CVNetBackbone=ResNet-101, Strategy=Global feature retrieval + Local feature reranking, Reranking Depth=4002023.08 | 87.2 | — | |
| SuperGlobalBackbone=ResNet-101, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 85.3 | — | |
| SuperGlobalBackbone=ResNet-101, Strategy=Global feature retrieval2023.08 | 85.3 | — | |
| SuperGlobalBackbone=ResNet-50, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 83.9 | — | |
| SuperGlobalBackbone=ResNet-50, Strategy=Global feature retrieval2023.08 | 83.9 | — | |
| DOLGBackbone=ResNet-101, Training Dataset=GLDv2, Result Source=Reproduced code2023.09 | 82.4 | — | |
| ViTGaLBackbone=XCiT-S24, Training Dataset=GLDv2, Result Source=Original paper [25]2023.09 | 82.4 | — | |
| DOLGBackbone=ResNet-101, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 81.5 | — | |
| DOLGBackbone=ResNet-101, Strategy=Global feature retrieval2023.08 | 81.5 | — | |
| DOLGBackbone=ResNet-101, Training Dataset=GLDv2, Result Source=Original paper [46]2023.09 | 81.5 | — | |
| DELGBackbone=ResNet-101, Retrieval Category=Global feature retrieval + Local feature reranking, Reranking Setting=GV rerank top 1002023.08 | 81.2 | — | |
| DELGBackbone=ResNet-101, Strategy=Global feature retrieval + Local feature reranking, Reranking Depth=1002023.08 | 81.2 | — | |
| CVNetBackbone=ResNet-50, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 81 | — | |
| CVNetBackbone=ResNet-50, Strategy=Global feature retrieval2023.08 | 81 | — | |
| RRT# Reranked images=400, Desc. version=R50-v2-clean2021.03 | 80.5 | — | |
| DOLGBackbone=ResNet-50, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 80.5 | — | |
| DOLGBackbone=ResNet-50, Strategy=Global feature retrieval2023.08 | 80.5 | — | |
| alphaQE + RRT# Reranked images=200, Desc. version=R50-v2-clean2021.03 | 80.4 | — | |
| CVNetBackbone=ResNet-101, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 80.2 | — | |
| CVNetBackbone=ResNet-101, Strategy=Global feature retrieval2023.08 | 80.2 | — | |
| ViTGaLBackbone=XCiT-S24, Training Dataset=GLDv2, Result Source=Reproduced code2023.09 | 79.6 | — | |
| RRT# Reranked images=200, Desc. version=R50-v2-clean2021.03 | 79.5 | — | |
| RRT# local desc.=500, # Reranked images=200, Desc. version=R50-v2-clean2021.03 | 79.5 | — | |
| RRT# Reranked images=400, Desc. version=R50-v12021.03 | 79.2 | — | |
| GV*# local desc.=1000, # Reranked images=200, Desc. version=R50-v2-clean2021.03 | 79.2 | — | |
| alphaQE + RRT# Reranked images=200, Desc. version=R50-v12021.03 | 78.7 | — | |
| GV# local desc.=1000, # Reranked images=100, Desc. version=R50-v2-clean2021.03 | 78.3 | — | |
| DELGBackbone=ResNet-50, Retrieval Category=Global feature retrieval + Local feature reranking, Reranking Setting=GV rerank top 1002023.08 | 78.3 | — | |
| DELGBackbone=ResNet-50, Strategy=Global feature retrieval + Local feature reranking, Reranking Depth=1002023.08 | 78.3 | — | |
| RRT# Reranked images=100, Desc. version=R50-v2-clean2021.03 | 78.1 | — | |
| RRT# local desc.=500, # Reranked images=100, Desc. version=R50-v2-clean2021.03 | 78.1 | — | |
| RRT# Reranked images=200, Desc. version=R50-v12021.03 | 77.9 | — | |
| RRT# local desc.=500, # Reranked images=200, Desc. version=R50-v12021.03 | 77.9 | — | |
| CiDeR-FTTraining Set=RGLDv2-clean, Fine-tuned=true2024.04 | 77.8 | 88 | |
| GV*# local desc.=1000, # Reranked images=200, Desc. version=R50-v12021.03 | 77.2 | — | |
| alphaQEDesc. version=R50-v2-clean2021.03 | 76.6 | — | |
| DELGBackbone=ResNet-101, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 76.3 | — | |
| DELGBackbone=ResNet-101, Strategy=Global feature retrieval2023.08 | 76.3 | — | |
| CiDeR-FTTrain Set=SfM-120k, Backbone=ResNet-101, Pooling=GeM, Loss=ArcFace, Fine-tuned=true, Dimension=20482024.04 | 76.2 | — | |
| CiDeR-FTTraining Set=SfM-120k, Fine-tuned=true2024.04 | 76.2 | 87.3 | |
| RRT# Reranked images=100, Desc. version=R50-v12021.03 | 75.5 | — | |
| RRT# local desc.=500, # Reranked images=100, Desc. version=R50-v12021.03 | 75.5 | — | |
| HOW N/DBackbone=ResNet-18, Paired day-night training=true2023.09 | 75.5 | — | |
| HOW N/DBackbone=ResNet-18, Paired day-night training data=true2023.09 | 75.5 | — | |
| GV# local desc.=1000, # Reranked images=100, Desc. version=R50-v12021.03 | 75.4 | — | |
| HOWBackbone=ResNet-18, Paired day-night training=false2023.09 | 75.1 | — | |
| HOWBackbone=ResNet-18, Paired day-night training data=false2023.09 | 75.1 | — | |
| RCFNGAN CDBackbone=ResNet-18, Paired day-night training=false2023.09 | 74.6 | — | |
| HEDNGAN CDBackbone=ResNet-18, Paired day-night training=false2023.09 | 74.6 | — | |
| CycleGAN CDBackbone=ResNet-18, Paired day-night training=false2023.09 | 74.6 | — | |
| HED^N GAN CDBackbone=ResNet-18, Paired day-night training data=false2023.09 | 74.6 | — | |
| CycleGAN CDBackbone=ResNet-18, Paired day-night training data=false2023.09 | 74.6 | — | |
| CiDeRTraining Set=RGLDv2-clean, Fine-tuned=false2024.04 | 73.7 | 85.5 | |
| DELG globalDesc. version=R50-v2-clean2021.03 | 73.6 | — | |
| DELG global# local desc.=0, # Reranked images=0, Desc. version=R50-v2-clean2021.03 | 73.6 | — | |
| DELGBackbone=ResNet-50, Retrieval Category=Global feature retrieval, Reranking Setting=No reranking2023.08 | 73.6 | — | |
| DELGBackbone=ResNet-50, Strategy=Global feature retrieval2023.08 | 73.6 | — | |
| alphaQEDesc. version=R50-v12021.03 | 72.9 | — | |
| GLAMTraining Set=RGLDv2-clean, Fine-tuned=false2024.04 | 72.2 | 84.7 | |
| R-ASMK*+SPTrain Set=NC-clean, Backbone=ResNet-50, Pooling=Softmax, Local, Fine-tuned=true2024.04 | 71.9 | — | |
| SOLARTraining Set=RGLDv2-clean, Fine-tuned=true2024.04 | 70.8 | 84.6 | |
| DOLGTraining Set=RGLDv2-clean, Fine-tuned=false2024.04 | 70.8 | 85.3 | |
| R-ASMK*Train Set=NC-clean, Backbone=ResNet-50, Pooling=Softmax, Local, Fine-tuned=true2024.04 | 69.9 | — | |
| DELG globalDesc. version=R50-v12021.03 | 69.7 | — | |
| DELG global# local desc.=0, # Reranked images=0, Desc. version=R50-v12021.03 | 69.7 | — | |
| DELF-ASMK*+SPTrain Set=SfM-120k, Backbone=ResNet-50, Pooling=Softmax, Fine-tuned=false2024.04 | 67.8 | — | |
| CiDeRTrain Set=SfM-120k, Backbone=ResNet-101, Pooling=GeM, Loss=ArcFace, Fine-tuned=false, Dimension=20482024.04 | 67.3 | — | |
| CiDeRTraining Set=SfM-120k, Fine-tuned=false2024.04 | 67.3 | 85.1 | |
| AGEMTrain Set=SfM-120k, Backbone=ResNet-101, Pooling=GeM, Loss=Siamese, Fine-tuned=true, Dimension=20482024.04 | 67 | — | |
| AGEMTraining Set=SfM-120k, Fine-tuned=true2024.04 | 67 | — | |
| CycleGAN CDBackbone=ResNet-101, Paired day-night training data=false2023.09 | 66.8 | — | |
| HED^N GAN CDBackbone=ResNet-101, Paired day-night training data=false2023.09 | 66.3 | — | |
| GLAMTrain Set=SfM-120k, Backbone=ResNet-101, Pooling=GeM, Loss=ArcFace, Fine-tuned=true, Dimension=5122024.04 | 66.2 | — | |
| GLAMTraining Set=SfM-120k, Fine-tuned=false2024.04 | 66.2 | — | |
| GeMBackbone=ResNet-101, Paired day-night training data=false2023.09 | 65.3 | — | |
| Radenovic et al.Train Set=SfM-120k, Backbone=ResNet-101, Pooling=GeM, Loss=Siamese, Fine-tuned=true, Dimension=20482024.04 | 64.7 | — | |
| Filip et al.Training Set=SfM-120k, Fine-tuned=true2024.04 | 64.7 | 84.7 | |
| Yokoo et al.Training Set=RGLDv2-clean, Fine-tuned=false2024.04 | 64.5 | 81 | |
| CIConvBackbone=ResNet-101, Paired day-night training data=false2023.09 | 62 | — | |
| HED^N GAN CDBackbone=VGG-16, Paired day-night training data=false2023.09 | 61.1 | — | |
| GeMBackbone=VGG-16, Paired day-night training data=false2023.09 | 60.9 | — | |
| TokenTraining Set=RGLDv2-clean, Fine-tuned=false2024.04 | 60.8 | 77.7 | |
| CLAHE CBackbone=VGG-16, Paired day-night training data=false2023.09 | 60.8 | — | |
| CycleGAN CDBackbone=VGG-16, Paired day-night training data=false2023.09 | 60.7 | — | |
| HesAff-rSIFT-ASMK*+SPTrain Set=SfM-120k, Backbone=ResNet-50, Fine-tuned=false2024.04 | 60.6 | — | |
| CLAHE N/D CBackbone=VGG-16, Paired day-night training data=true2023.09 | 60.2 | — | |
| GeM N/DBackbone=VGG-16, Paired day-night training data=true2023.09 | 60 | — | |
| GeMTraining Set=SfM-120k, Fine-tuned=false2024.04 | 54 | 72.5 | |
| DIRTrain Set=SfM-120k, Backbone=ResNet-101, Pooling=RMAC, Loss=Triplet, Fine-tuned=true, Dimension=20482024.04 | 53.5 | — |