Revisiting Oxford and Paris

Paper

Revisiting Oxford and Paris: Large-Scale Image Retrieval Benchmarking
Radenović F., Iscen A., Tolias G., Avrithis Y., Chum O.
CVPR 2018 [ pdf | bib ]

Oxford dataset images can be downloaded from the original webpage.
Paris dataset images can be downloaded from the original webpage.
R-1M distractor dataset images are available here.
Revisited annotations including new query image list (55 original and 15 additional query images), their corresponding bounding boxes, and database image list with excluded query images:

MATLAB structure:
- gnd_roxford5k.mat
- gnd_rparis6k.mat
Python dictionary:
- gnd_roxford5k.pkl
- gnd_rparis6k.pkl

Example code explaining how to load and process database and distractors images; load, crop and process query images; and evaluate three different protocols. Datasets, revisited annotations, and example features are automatically downloaded with example scripts. Available for:
- MATLAB
- Python