Datasets › ROxford

ROxford

Introduced by Filip Radenović et al. in Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking29 Mar 2018 archive 2025-07-28

@inproceedings{RITAC18,
 author = {Radenovi\'{c}, F. and Iscen, A. and Tolias, G. and Avrithis, Y. and Chum, O.},
 title = {Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking},
 booktitle = {CVPR},
 year = {2018}
}

Benchmarks archive 2025-07-28

All 3 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Image Retrieval ROxford (Hard) SuperGlobal mAP 80.2 Global Features are All You Need for Image Retrieval and... shihaoshao-gh/superglobal +1 23 Compare
Image Retrieval ROxford (Medium) AMES mAP 90.7 AMES: Asymmetric and Memory-Efficient Similarity... pavelsuma/ames 23 Compare
Image Retrieval ROxford Medium without fine-tuning HesAff–rSIFT–VLAD Average mAP 33.9 Revisiting Oxford and Paris: Large-Scale Image Retrieval... tensorflow/models +1 1 Compare

Papers archive 2025-07-28

17 shown of 17 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 73. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
AMES: Asymmetric and Memory-Efficient Similarity Estimation for Instance-level Retrieval 1 2 6 Aug 2024 ran 13 of 13 samples (0 unverified)
Dark Side Augmentation: Generating Diverse Night Examples for Metric Learning 1 1 28 Sep 2023 ran 2 of 3 samples (1 unverified)
Global Features are All You Need for Image Retrieval and Reranking 2 1 14 Aug 2023 ran 13 of 18 samples (5 unverified)
Learning Super-Features for Image Retrieval 1 2 31 Jan 2022 not harvested
Learning Token-based Representation for Image Retrieval 1 2 12 Dec 2021 not harvested
Hypergraph Propagation and Community Selection for Objects Retrieval 1 2 1 Dec 2021 not harvested
Emerging Properties in Self-Supervised Vision Transformers 32 2 29 Apr 2021 ran 5 of 20 samples (15 unverified; 2 pointer-only for licence)
Instance-level Image Retrieval using Reranking Transformers 1 2 22 Mar 2021 ran 2 of 3 samples (1 unverified; 3 pointer-only for licence)
Learning and aggregating deep local descriptors for instance-level recognition 2 2 26 Jul 2020 ran 1 of 3 samples (2 unverified)
Google Landmarks Dataset v2 -- A Large-Scale Benchmark for Instance-Level Recognition and Retrieval 5 2 3 Apr 2020 not harvested
Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking 2 15 29 Mar 2018 not harvested
Fine-tuning CNN Image Retrieval with No Human Annotation 14 2 3 Nov 2017 ran 3 of 24 samples (21 unverified; 2 pointer-only for licence)
Large-Scale Image Retrieval with Attentive Deep Local Features 13 4 19 Dec 2016 ran 3 of 12 samples (9 unverified)
End-to-end Learning of Deep Visual Representations for Image Retrieval 4 1 25 Oct 2016 not harvested
Cross-dimensional Weighting for Aggregated Deep Convolutional Features 1 2 13 Dec 2015 not harvested
Particular object retrieval with integral max-pooling of CNN activations 6 3 18 Nov 2015 ran 0 of 4 samples (4 unverified)
Aggregating Deep Convolutional Features for Image Retrieval 2 2 26 Oct 2015 not harvested

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • ROxford (Medium)
  • ROxford (Hard)
  • ROxford Medium without fine-tuning
  • ROxford

4 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections