Polarimetric HRRP Recognition Based on ConvLSTM With Self-Attention

Zhang, Liang; Li, Yang; Wang, Yanhua; Wang, Junfu; Li, Teng

doi:10.1109/jsen.2020.3044314

Cited by 30 publications

(7 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The MSGF‐1D‐CNN is configured with the following parameters: the window size and step size of the max‐pooling are both set to 3, and the stride of all convolution kernels is set to 1. In addition to comparing MSGF‐1D‐CNN with standard 1D‐CNN to verify the improvement, MSGF‐1D‐CNN is also compared with the stacked denoising sparse AE (sDSAE) [31], Bidirectional LSTM (Bi‐LSTM) [1], Bidirectional GRU (Bi‐GRU) [32], one‐dimensional local receptive field‐based extreme learning auto‐encoder (ELM‐LRF‐AE) [2], convolutional long short‐term memory (ConvLSTM) [33] and one‐dimensional convolutional neural network with channel attention (CNN1D‐CA) [15]. All the above DNN‐based HRRP recognition methods take advantage of the diffGrad algorithm [30] for training.…”

Section: Experiments and Discussionmentioning

confidence: 99%

Multi‐scale group‐fusion convolutional neural network for high‐resolution range profile target recognition

Xiang

Wang

Lai

et al. 2022

IET Radar Sonar & Navi

View full text Add to dashboard Cite

Convolution neural networks (CNNs) represent one of the workhorses of artificial intelligence applications. As a typical artificial intelligence application, a high-resolution range profile (HRRP) target recognition method based on CNNs has aroused a lot of research interest. Most CNNs use a relatively small and single-scale convolution kernel size to control the number of parameters and computational complexity, but recent studies indicate that CNNs with a small kernel size cannot extract enough spatial information, which hurts the recognition performance. Aiming at this problem, this paper proposes a multi-scale group-fusion one-dimensional convolution neural network (MSGF-1D-CNN) for HRRP target recognition. MSGF-1D-CNN utilises multi-scale group one-dimensional convolution (MSG 1D-Conv) and point-wise convolution (PW-Conv) to replace the standard convolution. Multi-scale group one-dimensional convolution can significantly reduce complexity and capture the information of targets within HRRP in different levels of detail to enhance feature extraction, while PW-Conv can realise the fusion of multi-scale features to help boosting recognition performance. Experiments on five mid-course ballistic targets in the HRRP dataset show that MSGF-1D-CNN has superior recognition performance, and the parameter number of the model is reduced by more than 2.4 times than standard 1D-CNN. Furthermore, MSGF-1D-CNN shows better performance on fine-grained HRRP target recognition and antinoise robustness in most cases. K E Y W O R D Sconvolution neural networks, feature fusion, high-resolution range profile, multi-scale group convolution, midcourse ballistic targets, point-wise convolutionThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

show abstract

Section: Experiments and Discussionmentioning

confidence: 99%

Multi‐scale group‐fusion convolutional neural network for high‐resolution range profile target recognition

Xiang

Wang

Lai

et al. 2022

IET Radar Sonar & Navi

View full text Add to dashboard Cite

show abstract

“…Commonly used sensitivity removal approaches include time-shift compensation, energy normalization, and average processing [54]- [56]. The DNN structures used for radar HRRP target recognition include the deep belief network [54], [55], recurrent attentional network [57], [58], concatenated neural network, CNNs [62]- [64], stacked auto-encoder (SAE) [65], and convolutional LSTM [66], [67].…”

Section: A Dl-based Atr Using Hrr Profilesmentioning

confidence: 99%

“…Some researchers used measured HRRP data for performance evaluation. For example, the HRRP data from Yak-42 (large jet), Cessna Citation S/II (small jet), and An-26 (twin-engine turboprop) were used in [54]- [58]; the HRRP data from Airbus A319, A320, A321, and Boeing B738 were used in [59]; the HRRP data from seven types of ship of different sizes (length from 89.3 m to 182.8 m) were used in [60]; the HRRP data from various types of ground vehicles were used in [62], [66], [67]. Since most researchers only have access to a limited mount of HRRP measurement data associated with a handful of vehicles, many of them resort to simulated HRRP data generated by software based on the specific CAD models of vehicles for research purposes.…”

Section: A Dl-based Atr Using Hrr Profilesmentioning

confidence: 99%

Deep-Learning for Radar: A Survey

Geng¹,

Zhang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

A comprehensive and well-structured review on the application of deep learning (DL) based algorithms, such as convolutional neural networks (CNN) and long-short term memory (LSTM), in radar signal processing is given. The following DL application areas are covered: i) radar waveform and antenna array design; ii) passive or low probability of interception (LPI) radar waveform recognition; iii) automatic target recognition (ATR) based on high range resolution profiles (HRRPs), Doppler signatures, and synthetic aperture radar (SAR) images; and iv) radar jamming/clutter recognition and suppression. Although DL is unanimously praised as the ultimate solution to many bottleneck problems in most of existing works on similar topics, both the positive and the negative sides of stories about DL are checked in this work. Specifically, two limiting factors of the real-life performance of deep neural networks (DNNs), limited training samples and adversarial examples, are thoroughly examined. By investigating the relationship between the DL-based algorithms proposed in various papers and linking them together to form a full picture, this work serves as a valuable source for researchers who are seeking potential research opportunities in this promising research field.

show abstract

“…The major concerns of RTR mainly relate to data acquisition, semantic feature discovery, and extraction methods. High-resolution signals like high-resolution range profile (HRRP) [1], [2], synthetic aperture radar (SAR) images [3], and inverse SAR (ISAR) images [4], [5] can present rich information [6] of targets but demand powerful radar. The radar cross section (RCS) signal, which characterizes the scattering shape and the movement pattern of the target, is widely used due to its easy availability and sufficient information for RTR.…”

Section: Introductionmentioning

confidence: 99%

Semantic Feature-Enhanced Graph ATtention Network for Radar Target Recognition in Heterogeneous Radar Network

et al. 2023

View full text Add to dashboard Cite

Radar target recognition (RTR), as a key technique of intelligent radar systems, has been widely investigated. Accurate RTR at low signal-to-noise ratios (SNRs) still remains an open challenge. Considering that most existing methods are based on a single radar or the homogeneous radar network, we extend RTR to the heterogeneous radar network to improve the robustness of RTR, which uses the radar cross section (RCS) signals at low SNRs by further exploiting the frequency-domain information. In this article, a Semantic Feature-Enhanced Graph ATtention Network (SFE-GAT) is proposed, which extracts semantic features from both the source and transform domains via the long short-term memory (LSTM) and GAT layers, then fuses them in the semantic space using an attention mechanism, and further distills higher-level semantic features using a GAT layer before classification. Extensive experiments are carried out to validate that the proposed SFE-GAT model can greatly improve the RTR accuracy in the low SNR region.

show abstract

Polarimetric HRRP Recognition Based on ConvLSTM With Self-Attention

Cited by 30 publications

References 61 publications

Multi‐scale group‐fusion convolutional neural network for high‐resolution range profile target recognition

Multi‐scale group‐fusion convolutional neural network for high‐resolution range profile target recognition

Deep-Learning for Radar: A Survey

Semantic Feature-Enhanced Graph ATtention Network for Radar Target Recognition in Heterogeneous Radar Network

Contact Info

Product

Resources

About