Authors:
Christoph Praschl
1
;
Roland Kaiser
2
and
Gerald Zwettler
1
;
3
Affiliations:
1
Research Group Advanced Information Systems and Technology, Research and Development Department, University of Applied Sciences Upper Austria, Softwarepark 11, Hagenberg, 4232, Austria
;
2
ENNACON Environment Nature Consulting KG, Altheim 13, Feldkirchen bei Mattighofen, Austria
;
3
Department of Software Engineering, School of Informatics, Communications and Media, University of Applied Sciences Upper Austria, Softwarepark 11, Hagenberg, 4232, Austria
Keyword(s):
Plant Cover Image Data Synthesis, Generative Adversial Networks, Deep Learning Instance Segmentation, Small Training Data Sets.
Abstract:
Deep learning approaches are highly influenced by two factors, namely the complexity of the task and the size of the training data set. In terms of both, the extraction of features of low-stature alpine plants represents a challenging domain due to their fuzzy appearance, a great structural variety in plant organs and the high effort associated with acquiring high-quality training data for such plants. For this reason, this study proposes an approach for training deep learning models in the context of alpine vegetation based on a combination of real-world and artificial data synthesised using Generative Adversarial Networks. The evaluation of this approach indicates that synthetic data can be used to increase the size of training data sets. With this at hand, the results and robustness of deep learning models are demonstrated with a U-Net segmentation model. The evaluation is carried out using a cross-validation for three alpine plants, namely Soldanella pusilla, Gnaphalium supinum,
and Euphrasia minima. Improved segmentation accuracy was achieved for the latter two species. Dice Scores of 24.16% vs 26.18% were quantified for Gnaphalium with 100 real-world training images. In the case of Euphrasia, Dice Scores improved from 33.56% to 42.96% using only 20 real-world training images.
(More)