Haitz, Dennis ; Hübner, Patrick ; Ulrich, Markus ; Landgraf, Steven ; Jutzi, Boris (2022)
Semantic Segmentation with Small Training Datasets: A Case Study for Corrosion Detection on the Surface of Industrial Objects.
Image Processing Forum 2022. Karlsruhe (24.11.2022-25.11.2022)
doi: 10.5445/IR/1000154095
Konferenzveröffentlichung, Bibliographie
Kurzbeschreibung (Abstract)
In this research, we investigate possibilities to train convolutional neural networks with a small dataset for semantic segmentation, while achieving the best possible model generalization. In particular, we want to segment corrosion on the surface of industrial objects. In order to achieve model generalization, we utilize a selection of established and advanced strategies, i.e. Self-Supervised-Learning. Besides radiometric- and geometric-based data augmentation, we focus on model complexity regarding encoder and decoder, as well as optimal pretraining. Finally, we evaluate the best performing model against a pixel-wise random forest classification. As a result, we achieve an f1-score of 0.79 for the best performing model regarding the segmentation of corrosion.
| Typ des Eintrags: | Konferenzveröffentlichung |
|---|---|
| Erschienen: | 2022 |
| Autor(en): | Haitz, Dennis ; Hübner, Patrick ; Ulrich, Markus ; Landgraf, Steven ; Jutzi, Boris |
| Art des Eintrags: | Bibliographie |
| Titel: | Semantic Segmentation with Small Training Datasets: A Case Study for Corrosion Detection on the Surface of Industrial Objects |
| Sprache: | Englisch |
| Publikationsjahr: | 2022 |
| Ort: | Karlsruhe, Germany |
| Verlag: | Karlsruher Institut für Technologie (KIT) |
| Titel der Zeitschrift, Zeitung oder Schriftenreihe: | Image Processing Forum |
| Buchtitel: | Forum Bildverarbeitung 2022 |
| Veranstaltungstitel: | Image Processing Forum 2022 |
| Veranstaltungsort: | Karlsruhe |
| Veranstaltungsdatum: | 24.11.2022-25.11.2022 |
| DOI: | 10.5445/IR/1000154095 |
| Kurzbeschreibung (Abstract): | In this research, we investigate possibilities to train convolutional neural networks with a small dataset for semantic segmentation, while achieving the best possible model generalization. In particular, we want to segment corrosion on the surface of industrial objects. In order to achieve model generalization, we utilize a selection of established and advanced strategies, i.e. Self-Supervised-Learning. Besides radiometric- and geometric-based data augmentation, we focus on model complexity regarding encoder and decoder, as well as optimal pretraining. Finally, we evaluate the best performing model against a pixel-wise random forest classification. As a result, we achieve an f1-score of 0.79 for the best performing model regarding the segmentation of corrosion. |
| Fachbereich(e)/-gebiet(e): | 13 Fachbereich Bau- und Umweltingenieurwissenschaften 13 Fachbereich Bau- und Umweltingenieurwissenschaften > Institut für Geodäsie 13 Fachbereich Bau- und Umweltingenieurwissenschaften > Institut für Geodäsie > Fernerkundung und Bildanalyse |
| Hinterlegungsdatum: | 27 Apr 2023 07:04 |
| Letzte Änderung: | 18 Okt 2024 11:51 |
| PPN: | 507907906 |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum