Pinpoint combines Flickr and street-view imagery via contrastive embedding and attention reranking to report state-of-the-art geolocation accuracy on IM2GPS3k, YFCC4k, and OSV-5M.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2verdicts
UNVERDICTED 2representative citing papers
The LIP-additive Asplund distance map of an image equals the LIP-multiplicative Asplund distance map of a transformed image, related by the LIP isomorphism.
citing papers explorer
-
Pinpoint: Grounded Worldwide Image Geolocation via Cross-Source Retrieval and Reranking
Pinpoint combines Flickr and street-view imagery via contrastive embedding and attention reranking to report state-of-the-art geolocation accuracy on IM2GPS3k, YFCC4k, and OSV-5M.
-
A Link Between the Multiplicative and Additive Functional Asplund's Metrics
The LIP-additive Asplund distance map of an image equals the LIP-multiplicative Asplund distance map of a transformed image, related by the LIP isomorphism.