Details
| Original language | English |
|---|---|
| Title of host publication | 2021 IEEE Statistical Signal Processing Workshop, SSP 2021 |
| Pages | 81-85 |
| Number of pages | 5 |
| ISBN (electronic) | 978-1-7281-5767-2 |
| Publication status | Published - 2021 |
Publication series
| Name | IEEE Workshop on Statistical Signal Processing Proceedings |
|---|---|
| Volume | 2021-July |
Abstract
Many modern applications rely on machine learning to fulfill their purpose. However, machine learning, especially the popular deep learning, requires a sufficient amount of labeled data to train models. For some tasks and in some domains, such as aerial images, labeling data is very time-consuming and thus expensive. We therefore propose strategies using unsupervised learning techniques to identify a subset of the input data which actually needs to be labeled by an expert in order to train a well-performing model. With our strategies, which involve less manual labeling effort, we were able to reduce the amount of training data required to 16%. At the same time, the model trained with this small subset achieved better semantic segmentation performance (average accuracy increase: 0.6%, average mIoU increase: 1.3%) for aerial images than a model trained with the full dataset.
Keywords
- Aerial Images, Deep Learning, Neural Networks, Semantic Segmentation, Semi-Supervised Learning
ASJC Scopus subject areas
- Mathematics(all)
- Applied Mathematics
- Computer Science(all)
- Signal Processing
- Engineering(all)
- Electrical and Electronic Engineering
- Computer Science(all)
- Computer Science Applications
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
2021 IEEE Statistical Signal Processing Workshop, SSP 2021. 2021. p. 81-85 9513774 (IEEE Workshop on Statistical Signal Processing Proceedings; Vol. 2021-July).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Minimizing Manual Labeling Effort for The Semantic Segmentation of Aerial Images
AU - Gritzner, Daniel
AU - Ostermann, Jörn
PY - 2021
Y1 - 2021
N2 - Many modern applications rely on machine learning to fulfill their purpose. However, machine learning, especially the popular deep learning, requires a sufficient amount of labeled data to train models. For some tasks and in some domains, such as aerial images, labeling data is very time-consuming and thus expensive. We therefore propose strategies using unsupervised learning techniques to identify a subset of the input data which actually needs to be labeled by an expert in order to train a well-performing model. With our strategies, which involve less manual labeling effort, we were able to reduce the amount of training data required to 16%. At the same time, the model trained with this small subset achieved better semantic segmentation performance (average accuracy increase: 0.6%, average mIoU increase: 1.3%) for aerial images than a model trained with the full dataset.
AB - Many modern applications rely on machine learning to fulfill their purpose. However, machine learning, especially the popular deep learning, requires a sufficient amount of labeled data to train models. For some tasks and in some domains, such as aerial images, labeling data is very time-consuming and thus expensive. We therefore propose strategies using unsupervised learning techniques to identify a subset of the input data which actually needs to be labeled by an expert in order to train a well-performing model. With our strategies, which involve less manual labeling effort, we were able to reduce the amount of training data required to 16%. At the same time, the model trained with this small subset achieved better semantic segmentation performance (average accuracy increase: 0.6%, average mIoU increase: 1.3%) for aerial images than a model trained with the full dataset.
KW - Aerial Images
KW - Deep Learning
KW - Neural Networks
KW - Semantic Segmentation
KW - Semi-Supervised Learning
UR - http://www.scopus.com/inward/record.url?scp=85113526781&partnerID=8YFLogxK
U2 - 10.1109/SSP49050.2021.9513774
DO - 10.1109/SSP49050.2021.9513774
M3 - Conference contribution
SN - 978-1-7281-5768-9
T3 - IEEE Workshop on Statistical Signal Processing Proceedings
SP - 81
EP - 85
BT - 2021 IEEE Statistical Signal Processing Workshop, SSP 2021
ER -