Please use this identifier to cite or link to this item: http://localhost/handle/Hannan/228211
Title: Zero-Shot Learning of SAR Target Feature Space With Deep Generative Neural Networks
Authors: Qian Song;Feng Xu
Year: 2017
Publisher: IEEE
Abstract: Zero-shot learning (ZSL) is of critical importance for practical synthetic aperture radar (SAR) automatic target recognition (ATR) as training samples are not always available for all targets and all observation configurations. We propose a novel generative-based deep neural network framework for ZSL of SAR ATR. The key component of the framework is a generative deconvolutional neural network referred to as generator. It learns a faithful hierarchical representation of known targets while automatically constructing a continuous SAR target feature space spanned by orientation-invariant features and orientation angle. It is then used as a reference to design and initialize an interpreter convolutional neural network, which is inversely symmetric to the generator network. The interpreter network is then trained to map any input SAR image, including those of unseen targets, into the target feature space. In a preliminary experiment with the Moving and Stationary Target Acquisition and Recognition data set, seven targets are used in the training of generator and interpreter networks. Then, the eighth target is used to test the interpreter, where it is correctly mapped to the reasonable spot spanned by the previous seven targets and its orientation can also be estimated.
URI: http://localhost/handle/Hannan/228211
volume: 14
issue: 12
More Information: 2245,
2249
Appears in Collections:2017

Files in This Item:
File SizeFormat 
8103339.pdf1.51 MBAdobe PDF
Title: Zero-Shot Learning of SAR Target Feature Space With Deep Generative Neural Networks
Authors: Qian Song;Feng Xu
Year: 2017
Publisher: IEEE
Abstract: Zero-shot learning (ZSL) is of critical importance for practical synthetic aperture radar (SAR) automatic target recognition (ATR) as training samples are not always available for all targets and all observation configurations. We propose a novel generative-based deep neural network framework for ZSL of SAR ATR. The key component of the framework is a generative deconvolutional neural network referred to as generator. It learns a faithful hierarchical representation of known targets while automatically constructing a continuous SAR target feature space spanned by orientation-invariant features and orientation angle. It is then used as a reference to design and initialize an interpreter convolutional neural network, which is inversely symmetric to the generator network. The interpreter network is then trained to map any input SAR image, including those of unseen targets, into the target feature space. In a preliminary experiment with the Moving and Stationary Target Acquisition and Recognition data set, seven targets are used in the training of generator and interpreter networks. Then, the eighth target is used to test the interpreter, where it is correctly mapped to the reasonable spot spanned by the previous seven targets and its orientation can also be estimated.
URI: http://localhost/handle/Hannan/228211
volume: 14
issue: 12
More Information: 2245,
2249
Appears in Collections:2017

Files in This Item:
File SizeFormat 
8103339.pdf1.51 MBAdobe PDF
Title: Zero-Shot Learning of SAR Target Feature Space With Deep Generative Neural Networks
Authors: Qian Song;Feng Xu
Year: 2017
Publisher: IEEE
Abstract: Zero-shot learning (ZSL) is of critical importance for practical synthetic aperture radar (SAR) automatic target recognition (ATR) as training samples are not always available for all targets and all observation configurations. We propose a novel generative-based deep neural network framework for ZSL of SAR ATR. The key component of the framework is a generative deconvolutional neural network referred to as generator. It learns a faithful hierarchical representation of known targets while automatically constructing a continuous SAR target feature space spanned by orientation-invariant features and orientation angle. It is then used as a reference to design and initialize an interpreter convolutional neural network, which is inversely symmetric to the generator network. The interpreter network is then trained to map any input SAR image, including those of unseen targets, into the target feature space. In a preliminary experiment with the Moving and Stationary Target Acquisition and Recognition data set, seven targets are used in the training of generator and interpreter networks. Then, the eighth target is used to test the interpreter, where it is correctly mapped to the reasonable spot spanned by the previous seven targets and its orientation can also be estimated.
URI: http://localhost/handle/Hannan/228211
volume: 14
issue: 12
More Information: 2245,
2249
Appears in Collections:2017

Files in This Item:
File SizeFormat 
8103339.pdf1.51 MBAdobe PDF