TL;DR
This paper investigates whether large-scale, noisy, publicly available map data can effectively train CNNs for aerial image segmentation, reducing manual annotation efforts while maintaining high performance.
Contribution
It demonstrates that training CNNs with noisy, large-scale data from sources like OpenStreetMap can achieve competitive segmentation results, lessening the need for manual labeling.
Findings
Noisy large-scale data can be effectively used for training.
Comparable performance achieved with less manual annotation.
OpenStreetMap data can substitute manual labels in some cases.
Abstract
This study deals with semantic segmentation of high-resolution (aerial) images where a semantic class label is assigned to each pixel via supervised classification as a basis for automatic map generation. Recently, deep convolutional neural networks (CNNs) have shown impressive performance and have quickly become the de-facto standard for semantic segmentation, with the added benefit that task-specific feature design is no longer necessary. However, a major downside of deep learning methods is that they are extremely data-hungry, thus aggravating the perennial bottleneck of supervised classification, to obtain enough annotated training data. On the other hand, it has been observed that they are rather robust against noise in the training labels. This opens up the intriguing possibility to avoid annotating huge amounts of training data, and instead train the classifier from existing…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
