Classifying figures and illustrations in electronics datasheets: A comparative evaluation of recent computer vision models on a custom collection of 4000 technical documents
dc.contributor.author | Perakis, Lymperis | |
dc.contributor.author | Balling, Julian | |
dc.contributor.author | Binder, Frank | |
dc.contributor.author | Heyer, Gerhard | |
dc.contributor.author | Kreupl, Franz | |
dc.contributor.editor | Klein, Maike | |
dc.contributor.editor | Krupka, Daniel | |
dc.contributor.editor | Winter, Cornelia | |
dc.contributor.editor | Wohlgemuth, Volker | |
dc.date.accessioned | 2023-11-29T14:50:24Z | |
dc.date.available | 2023-11-29T14:50:24Z | |
dc.date.issued | 2023 | |
dc.description.abstract | We report findings from a comparative evaluation of several recent object detection models applied to a domain-specific use case in technical document analysis and graphics recognition. More specifically, we apply models from the EfficientDet and YOLO model families to detect and classify figures in electronics datasheets according to a custom classification scheme. We identify YOLOv7-D6 as the most accurate model in our study and show that it can successfully solve this task. We highlight an iterative approach to figure annotation in document page images for creating a comprehensive and balanced custom dataset for our use case. In our experiments, the object detection models show impressive performance levels on par with state-of-the-art results from the literature and related studies. | en |
dc.identifier.doi | 10.18420/inf2023_186 | |
dc.identifier.isbn | 978-3-88579-731-9 | |
dc.identifier.pissn | 1617-5468 | |
dc.identifier.uri | https://dl.gi.de/handle/20.500.12116/43114 | |
dc.language.iso | en | |
dc.publisher | Gesellschaft für Informatik e.V. | |
dc.relation.ispartof | INFORMATIK 2023 - Designing Futures: Zukünfte gestalten | |
dc.relation.ispartofseries | Lecture Notes in Informatics (LNI) - Proceedings, Volume P-337 | |
dc.subject | Computer Vision | |
dc.subject | Object Detection | |
dc.subject | Document Analysis | |
dc.subject | Graphics Recognition | |
dc.subject | Electronic Design Automation | |
dc.subject | Machine Learning | |
dc.title | Classifying figures and illustrations in electronics datasheets: A comparative evaluation of recent computer vision models on a custom collection of 4000 technical documents | en |
dc.type | Text/Conference Paper | |
gi.citation.endPage | 1848 | |
gi.citation.publisherPlace | Bonn | |
gi.citation.startPage | 1833 | |
gi.conference.date | 26.-29. September 2023 | |
gi.conference.location | Berlin | |
gi.conference.sessiontitle | Wirtschaft, Management Industrie - Künstliche Intelligenz für kleine und mittlere Unternehmen (KI-KMU 2023) |
Dateien
Originalbündel
1 - 1 von 1
Vorschaubild nicht verfügbar
- Name:
- 09_02_03_Perakis.pdf
- Größe:
- 402.52 KB
- Format:
- Adobe Portable Document Format