TY - DATA
T1 - AmsterTime: A Visual Place Recognition Benchmark Dataset for Severe Domain Shift
PY - 2023/10/09
AU - Burak Yildiz
AU - Seyran Khademi
AU - Ronald Maria Siebes
AU - Jan van Gemert
AU - Tino Mager
AU - Beate Löffler
AU - Carola Hein
AU - Victor de Boer
UR - 
DO - 10.4121/19580806.v4
KW - Amsterdam
KW - Visual Place Recognition
KW - Benchmark dataset
KW - Mapillary
KW - Archimedial
KW - Archival images
KW - Street-view images
N2 - <p><strong>AmsterTime</strong> dataset offers a collection of 2,500 well-curated images matching the same scene from a street view matched to historical archival image data from Amsterdam city. The image pairs capture the same place with different cameras, viewpoints, and appearances. Unlike existing benchmark datasets, AmsterTime is directly crowdsourced in a GIS navigation platform (Mapillary). In turn, all the matching pairs are verified by a human expert to verify the correct matches and evaluate the human competence in the Visual Place Recognition (VPR) task for further references.</p><p><br></p><p>The properties of the dataset are summarized as:</p><ul><li>1200+ license-free images from the Amsterdam City Archive, representing urban places in the city of Amsterdam, captured in the past century by many photographers.</li><li>All archival queries are matched with street view images from Mapillary.</li><li>All matches are verified by architectural historians and Amsterdam inhabitants.</li><li>Image pairs are archival and street views capturing the same place with different cameras, time lags, structural changes, occlusion, viewpoint, appearance, and illuminations.</li><li>The dataset exhibits a domain shift between query and the gallery due to significant difference between scanned archival and street view images.</li></ul><p>Two sub-tasks are created on the dataset:</p><ul><li><strong>Verification</strong> is a binary classification (auxiliary) task to detect a pair of archival and street-view images of the same place. The verification task for AmsterTime dataset has all of the crowdsourced image pairs as positive labeled, where the same number of negative samples are generated by randomly pairing archival and street-view images summing up to a total of 2,462 pairs in the verification task.</li><li><strong>Retrieval</strong> is the main task corresponding to VPR, in which a given query image is matched with a set of gallery images. For the retrieval task, AmsterTime dataset offers 1231 query images where the leave-one-out set serves as the gallery images for each query.</li></ul>
ER -