Details
| Original language | English |
|---|---|
| Title of host publication | Computer Vision – ECCV 2018 |
| Subtitle of host publication | 15th European Conference, 2018, Proceedings |
| Editors | Martial Hebert, Vittorio Ferrari, Cristian Sminchisescu, Yair Weiss |
| Publisher | Springer Verlag |
| Pages | 575-592 |
| Number of pages | 18 |
| ISBN (print) | 9783030012571 |
| Publication status | Published - 6 Oct 2018 |
| Event | 15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany Duration: 8 Sept 2018 → 14 Sept 2018 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 11216 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 1611-3349 |
Abstract
While the successful estimation of a photo’s geolocation enables a number of interesting applications, it is also a very challenging task. Due to the complexity of the problem, most existing approaches are restricted to specific areas, imagery, or worldwide landmarks. Only a few proposals predict GPS coordinates without any limitations. In this paper, we introduce several deep learning methods, which pursue the latter approach and treat geolocalization as a classification problem where the earth is subdivided into geographical cells. We propose to exploit hierarchical knowledge of multiple partitionings and additionally extract and take the photo’s scene content into account, i.e., indoor, natural, or urban setting etc. As a result, contextual information at different spatial resolutions as well as more specific features for various environmental settings are incorporated in the learning process of the convolutional neural network. Experimental results on two benchmarks demonstrate the effectiveness of our approach outperforming the state of the art while using a significant lower number of training images and without relying on retrieval methods that require an appropriate reference dataset.
Keywords
- Context-based classification, Deep learning, Geolocation estimation, Scene classification
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
- General Computer Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Computer Vision – ECCV 2018: 15th European Conference, 2018, Proceedings. ed. / Martial Hebert; Vittorio Ferrari; Cristian Sminchisescu; Yair Weiss. Springer Verlag, 2018. p. 575-592 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11216 LNCS).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Geolocation estimation of photos using a hierarchical model and scene classification
AU - Müller-Budack, Eric
AU - Pustu-Iren, Kader
AU - Ewerth, Ralph
N1 - Funding information: Acknowledgement. This work is financially supported by the German Research Foundation (DFG: Deutsche Forschungsgemeinschaft, project number: EW 134/4-1).
PY - 2018/10/6
Y1 - 2018/10/6
N2 - While the successful estimation of a photo’s geolocation enables a number of interesting applications, it is also a very challenging task. Due to the complexity of the problem, most existing approaches are restricted to specific areas, imagery, or worldwide landmarks. Only a few proposals predict GPS coordinates without any limitations. In this paper, we introduce several deep learning methods, which pursue the latter approach and treat geolocalization as a classification problem where the earth is subdivided into geographical cells. We propose to exploit hierarchical knowledge of multiple partitionings and additionally extract and take the photo’s scene content into account, i.e., indoor, natural, or urban setting etc. As a result, contextual information at different spatial resolutions as well as more specific features for various environmental settings are incorporated in the learning process of the convolutional neural network. Experimental results on two benchmarks demonstrate the effectiveness of our approach outperforming the state of the art while using a significant lower number of training images and without relying on retrieval methods that require an appropriate reference dataset.
AB - While the successful estimation of a photo’s geolocation enables a number of interesting applications, it is also a very challenging task. Due to the complexity of the problem, most existing approaches are restricted to specific areas, imagery, or worldwide landmarks. Only a few proposals predict GPS coordinates without any limitations. In this paper, we introduce several deep learning methods, which pursue the latter approach and treat geolocalization as a classification problem where the earth is subdivided into geographical cells. We propose to exploit hierarchical knowledge of multiple partitionings and additionally extract and take the photo’s scene content into account, i.e., indoor, natural, or urban setting etc. As a result, contextual information at different spatial resolutions as well as more specific features for various environmental settings are incorporated in the learning process of the convolutional neural network. Experimental results on two benchmarks demonstrate the effectiveness of our approach outperforming the state of the art while using a significant lower number of training images and without relying on retrieval methods that require an appropriate reference dataset.
KW - Context-based classification
KW - Deep learning
KW - Geolocation estimation
KW - Scene classification
UR - http://www.scopus.com/inward/record.url?scp=85055120639&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-01258-8_35
DO - 10.1007/978-3-030-01258-8_35
M3 - Conference contribution
AN - SCOPUS:85055120639
SN - 9783030012571
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 575
EP - 592
BT - Computer Vision – ECCV 2018
A2 - Hebert, Martial
A2 - Ferrari, Vittorio
A2 - Sminchisescu, Cristian
A2 - Weiss, Yair
PB - Springer Verlag
T2 - 15th European Conference on Computer Vision, ECCV 2018
Y2 - 8 September 2018 through 14 September 2018
ER -