AmsterTime: A Visual Place Recognition Benchmark Dataset for Severe Domain Shift
CC0
AmsterTime: a challenging dataset to benchmark visual place recognition (VPR) in presence of a severe domain shift. AmsterTime 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). Various baselines, including non-learning, supervised and self-supervised methods, pre-trained on different relevant datasets, for both verification and retrieval tasks are evaluated. The result credits the best accuracy to the ResNet-101 model pre-trained on the Landmarks dataset for both verification and retrieval tasks by 84% and 24%, respectively. Additionally, a subset of Amsterdam landmarks is collected for feature evaluation in a classification task. Classification labels are further used to extract the visual explanations using Grad-CAM for inspection of the learned similar visuals in a deep metric learning models.
- 2022-04-12 first online, published, posted
- ArchiMediaL: Developing Post-colonial Interpretations of Built Form through Heterogeneous Linked Digital Media (grant code 91913) [more info...] Volkswagen Foundation
DATA
- 913,787,908 bytesMD5:
21b59f6a17ef115889dbbae4f7c21fccAmsterTime.zip -
download all files (zip)
913,787,908 bytes unzipped
