Multi-Task Learning for Segmentation of Building Footprints with Deep Neural Networks
By Benjamin Bischke, Patrick Helber, Joachim Folz, Damian Borth, Andreas R. Dengel. 2017
272 citations recorded by the source corpus.
The increased availability of high-resolution satellite imagery allows to sense very detailed structures on the surface of our planet. Access to such information opens up new directions in the analysis of remote sensing imagery. While deep neural networks have achieved significant advances in semantic segmentation of high-resolution images, most of the existing approaches tend to produce predictions with poor boundaries. In this paper, we address the problem of preserving semantic segmentation boundaries in high-resolution satellite imagery by introducing a novel multi-task loss. The loss leverages multiple output representations of the segmentation mask and biases the network to focus more on pixels near boundaries. We evaluate our approach on the large-scale Inria Aerial Image Labeling Dataset which contains high-resolution images. Our results show that we are able to outperform state-of-the-art methods by 9.8% on the Intersection over Union (IoU) metric without any additional post-processing steps. Source code and all models will be available under https://github.com/bbischke/MultiTaskBuildingSegmentation.
Subfields: Computer Vision and Pattern Recognition, Media Technology, Ocean Engineering, Remote Sensing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Xilin Chen
- Jingdong Wang
- Yuhui Yuan
- Luc Van Gool
- Alexey A. Shvets
- Vladimir I. Iglovikov
- Dacheng Tao
- Jingyi Xie
In Media Technology
- Alexey A. Shvets
- Vladimir I. Iglovikov
- Jingdong Wang
- Jingyi Xie
- Senthil Kumar Yogamani
- Ghassan Hamarneh
- Xilin Chen
- Yuhui Yuan
In Ocean Engineering
In Remote Sensing
Citations in the corpus
- Visual SLAM for Automated Driving: Exploring the Applications of Deep Learning (2018, background): “Interestingly, [11] uses a ranking loss that penalizes the non relative correspondence between predictions and ground truth while [4] defines depth estimation as a classification problem.”
- Deep semantic segmentation of natural and medical images: a review (2019, background): “ultiple tasks are learned simultaneously, and the learning efficiency and the model performance on each of the tasks are improved because of the existing commonalities across the tasks. Bischke et al. [12] proposed a cascaded multi-task loss to preserve boundary information from segmentation masks fo”
- TernausNet: U-Net with VGG11 Encoder Pre-Trained on ImageNet for Image Segmentation (2018, methodology): “We used 30 images (5 from every 6 cities in the train set) for validation, as suggested in [12] (valid.”
- SegFix: Model-Agnostic Boundary Refinement for Segmentation (2020, background): “The other work [16,6,10,48] proposed to regularize the semantic or instance segmentation predictions with distance map or direction map in a multi-task mechanism.”