Title
Street View Challenge: Identification Of Commercial Entities In Street View Imagery
Keywords
Commercial Entity; Store Front; Street View; Street View Challenge
Abstract
This paper presents our submission to the Street View Challenge of identifying commercial entities in street view imagery. The provided data set of the challenge consists of approximately 129K street view images tagged with GPScoordinates. The problem is to identify different types of businesses visible in these images. Our solution is based on utilizing the textual information. However, the textual content of street view images is challenging in terms of variety and complexity, which limits the success of the approaches that are purely based on processing the content. Therefore, we use a method which leverages both the textual content of the images and business listings, in order to accomplish the identification task successfully. The robustness of our method is due to the fact that the information obtained from the different resources is cross-validated leading to significant improvements compared to the baselines. The experiments show approximately 70% of success rate on the defined problem. © 2011 IEEE.
Publication Date
12-1-2011
Publication Title
Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011
Volume
2
Number of Pages
380-383
Document Type
Article; Proceedings Paper
Personal Identifier
scopus
DOI Link
https://doi.org/10.1109/ICMLA.2011.181
Copyright Status
Unknown
Socpus ID
84857805357 (Scopus)
Source API URL
https://api.elsevier.com/content/abstract/scopus_id/84857805357
STARS Citation
Zamir, Amir Roshan; Darino, Alexander; and Shah, Mubarak, "Street View Challenge: Identification Of Commercial Entities In Street View Imagery" (2011). Scopus Export 2010-2014. 2232.
https://stars.library.ucf.edu/scopus2010/2232