Deep Fusion of Imaging Modalities for Semantic Segmentation of Satellite Imagery
@inproceedings{Sundelius2018DeepFO,
title={Deep Fusion of Imaging Modalities for Semantic Segmentation of Satellite Imagery},
author={Carl Sundelius},
year={2018}
}In this report I summarize my master’s thesis work, in which I have investigated different approaches for fusing imaging modalities for semantic segmentation with deep convolutional networks. State ...
Figures and Tables from this paper
figure 1.1 
figure 2.1 
table 2.1 
figure 2.2 
table 2.2 
figure 2.3 
figure 2.4 
figure 2.5 
figure 2.6 
figure 2.7 
figure 2.8 
figure 2.9 
figure 3.1 
figure 3.2 
figure 3.3 
figure 3.4 
figure 3.5 
figure 4.1 
table 4.1 
figure 4.2 
table 4.2 
figure 4.3 
table 4.3 
figure 4.4 
table 4.4 
figure 4.5 
table 4.5 
figure 4.6 
table 4.6 
figure 4.7 
table 5.1 
figure 5.1 
figure 5.10 
figure 5.11 
table 5.2 
figure 5.2 
table 5.3 
figure 5.3 
table 5.4 
figure 5.4 
table 5.5 
figure 5.5 
table 5.6 
figure 5.6 
table 5.7 
figure 5.7 
table 5.8 
figure 5.8 
figure 5.9 
figure A.1 
figure A.2 
figure A.3 
figure A.4 
table B.1 
table B.10 
table B.11 
table B.12 
table B.13 
table B.14 
table B.15 
table B.2 
table B.3 
table B.4 
table B.5 
table B.6 
table B.7 
table B.8 
table B.9 
figure C.1 
figure C.1 
figure C.1 
figure C.3 
figure C.3 
figure C.3 
figure C.5 
figure C.5 
figure C.5 
figure C.7 
figure C.7 
figure C.7
One Citation
Global Semantic Land Use/Land Cover Based on High Resolution Satellite Imagery Using Ensemble Networks
- Environmental Science, Computer ScienceIGARSS 2020 - 2020 IEEE International Geoscience and Remote Sensing Symposium
- 2020
This paper describes an architecture for and the results of global Land Use/Land Cover semantic segmentation. It is based on Vricon 3D Surface Models and high resolution, better than 0.5m resolution,…
References
SHOWING 1-10 OF 29 REFERENCES
Gated Convolutional Neural Network for Semantic Segmentation in High-Resolution Images
- Computer ScienceRemote. Sens.
- 2017
A gated convolutional neural network is proposed that achieves competitive segmentation accuracy on the public ISPRS 2D Semantic Labeling benchmark, which is challenging for segmentation by only using the RGB images.
Deep Neural Networks for Semantic Segmentation of Multispectral Remote Sensing Imagery
- Computer Science, Environmental ScienceArXiv
- 2017
Two deep neural networks for semantic segmentation of multispectral remote sensing imagery are developed and it is shown that this significantly improves results on real-worldRemote sensing imagery, and a new state-of-the-art result on the challenging Hamlin Beach State Park Dataset is established.
SEMANTIC SEGMENTATION OF AERIAL IMAGES WITH AN ENSEMBLE OF CNNS
- Computer Science
- 2016
A FCN is designed which takes as input intensity and range data and, with the help of aggressive deconvolution and recycling of early network layers, converts them into a pixelwise classification at full resolution, and discusses design choices and intricacies of such a network.
Fully Convolutional Networks for Dense Semantic Labelling of High-Resolution Aerial Imagery
- Environmental Science, Computer ScienceArXiv
- 2016
To make better use of image features, a pre-trained CNN is fine-tuned on remote sensing data in a hybrid network context, resulting in superior results compared to a network trained from scratch.
Effective semantic pixel labelling with convolutional networks and Conditional Random Fields
- Computer Science2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
- 2015
An effective semantic pixel labelling using CNN features, hand-crafted features and Conditional Random Fields (CRFs) is proposed and applied to the ISPRS 2D semantic labelling challenge dataset with competitive classification accuracy.
Deep Deconvolutional Networks for Scene Parsing
- Computer ScienceArXiv
- 2014
A novel network architecture is proposed that combines deep deconvolutional neural networks with CNNs and is employed for multi-patch training, which has the added advantage of having a training system that can be completely automated end-to-end with- out requiring any post-processing.
Rethinking Atrous Convolution for Semantic Image Segmentation
- Computer ScienceArXiv
- 2017
The proposed `DeepLabv3' system significantly improves over the previous DeepLab versions without DenseCRF post-processing and attains comparable performance with other state-of-art models on the PASCAL VOC 2012 semantic image segmentation benchmark.
Improving Fully Convolution Network for Semantic Segmentation
- Computer ScienceArXiv
- 2016
The Improved Fully Convolution Network (IFCN) is presented, which introduces a context network that progressively expands the receptive fields of feature maps and dense skip connections are added so that the context network can be effectively optimized.
Multimodal deep learning for robust RGB-D object recognition
- Computer Science2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
- 2015
This paper leverages recent progress on Convolutional Neural Networks (CNNs) and proposes a novel RGB-D architecture for object recognition that is composed of two separate CNN processing streams - one for each modality - which are consecutively combined with a late fusion network.
Pyramid Scene Parsing Network
- Computer Science2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- 2017
This paper exploits the capability of global context information by different-region-based context aggregation through the pyramid pooling module together with the proposed pyramid scene parsing network (PSPNet) to produce good quality results on the scene parsing task.