Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Linking Theory and Practice of Digital Libraries |
| Herausgeber/-innen | Omar Alonso, Helena Cousijn, Gianmaria Silvello, Stefano Marchesin, Mónica Marrero, Carla Teixeira Lopes |
| Seiten | 182-191 |
| Seitenumfang | 10 |
| ISBN (elektronisch) | 978-3-031-43849-3 |
| Publikationsstatus | Veröffentlicht - 2023 |
Publikationsreihe
| Name | Lecture Notes in Computer Sciences |
|---|---|
| Band | 14241 |
| ISSN (Print) | 0302-9743 |
| ISSN (elektronisch) | 1611-3349 |
Abstract
Due to the swift growth of patent applications each year, information and multimedia retrieval approaches that facilitate patent exploration and retrieval are of utmost importance. Different types of visualizations (e.g., graphs, technical drawings) and perspectives (e.g., side view, perspective) are used to visualize details of innovations in patents. The classification of these images enables a more efficient search in digital libraries and allows for further analysis. So far, datasets for image type classification miss some important visualization types for patents. Furthermore, related work does not make use of recent deep learning approaches including transformers. In this paper, we adopt state-of-the-art deep learning methods for the classification of visualization types and perspectives in patent images. We extend the CLEF-IP dataset for image type classification in patents to ten classes and provide manual ground truth annotations. In addition, we derive a set of hierarchical classes from a dataset that provides weakly-labeled data for image perspectives. Experimental results have demonstrated the feasibility of the proposed approaches. Source code, models, and datasets are publicly available (https://github.com/TIBHannover/PatentImageClassification ).
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
- Allgemeine Computerwissenschaft
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Linking Theory and Practice of Digital Libraries. Hrsg. / Omar Alonso; Helena Cousijn; Gianmaria Silvello; Stefano Marchesin; Mónica Marrero; Carla Teixeira Lopes. 2023. S. 182-191 (Lecture Notes in Computer Sciences; Band 14241).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Classification of Visualization Types and Perspectives in Patents
AU - Ghauri, Junaid Ahmed
AU - Müller-Budack, Eric
AU - Ewerth, Ralph
N1 - Funding Information: Acknowledgment. Part of this work is financially supported by the BMBF (Federal Ministry of Education and Research, Germany) project “ExpResViP” (project no: 01IO2004A). We also like to thank our colleague Sushil Awale (TIB) for his valuable feedback.
PY - 2023
Y1 - 2023
N2 - Due to the swift growth of patent applications each year, information and multimedia retrieval approaches that facilitate patent exploration and retrieval are of utmost importance. Different types of visualizations (e.g., graphs, technical drawings) and perspectives (e.g., side view, perspective) are used to visualize details of innovations in patents. The classification of these images enables a more efficient search in digital libraries and allows for further analysis. So far, datasets for image type classification miss some important visualization types for patents. Furthermore, related work does not make use of recent deep learning approaches including transformers. In this paper, we adopt state-of-the-art deep learning methods for the classification of visualization types and perspectives in patent images. We extend the CLEF-IP dataset for image type classification in patents to ten classes and provide manual ground truth annotations. In addition, we derive a set of hierarchical classes from a dataset that provides weakly-labeled data for image perspectives. Experimental results have demonstrated the feasibility of the proposed approaches. Source code, models, and datasets are publicly available (https://github.com/TIBHannover/PatentImageClassification ).
AB - Due to the swift growth of patent applications each year, information and multimedia retrieval approaches that facilitate patent exploration and retrieval are of utmost importance. Different types of visualizations (e.g., graphs, technical drawings) and perspectives (e.g., side view, perspective) are used to visualize details of innovations in patents. The classification of these images enables a more efficient search in digital libraries and allows for further analysis. So far, datasets for image type classification miss some important visualization types for patents. Furthermore, related work does not make use of recent deep learning approaches including transformers. In this paper, we adopt state-of-the-art deep learning methods for the classification of visualization types and perspectives in patent images. We extend the CLEF-IP dataset for image type classification in patents to ten classes and provide manual ground truth annotations. In addition, we derive a set of hierarchical classes from a dataset that provides weakly-labeled data for image perspectives. Experimental results have demonstrated the feasibility of the proposed approaches. Source code, models, and datasets are publicly available (https://github.com/TIBHannover/PatentImageClassification ).
KW - deep learning
KW - digital libraries
KW - patent image classification
UR - http://www.scopus.com/inward/record.url?scp=85174608481&partnerID=8YFLogxK
U2 - 10.48550/arXiv.2307.10471
DO - 10.48550/arXiv.2307.10471
M3 - Conference contribution
SN - 978-3-031-43848-6
T3 - Lecture Notes in Computer Sciences
SP - 182
EP - 191
BT - Linking Theory and Practice of Digital Libraries
A2 - Alonso, Omar
A2 - Cousijn, Helena
A2 - Silvello, Gianmaria
A2 - Marchesin, Stefano
A2 - Marrero, Mónica
A2 - Teixeira Lopes, Carla
ER -