Search

The spatial-temporal sequence prediction task is a challenge for neural networks, as it requires capturing temporal and spatial changes simultaneously. Recent researches focus on the modification of the ConvLSTM internal elements, which improves the prediction ability to some extent but introduces a large number of parameters. We observe that although the result generated by the row ConvLSTM is not very accurate in position and have a blurry appearance, it contains enough elements required to reconstruct the prediction. The motivation is to further refine these feature maps. Therefore, we propose a multi-attention LSTM (MA-LSTM) based on dimensional decoupling attention to alleviate the problems of the existing ConvLSTM-based methods, including inaccurate positions prediction and blurry generated results. The proposed model includes the Dimensionality Decoupling Module (D2M) and the Channel Attention Module (CAM). D2M compresses dimensions to transfer motion features over time that squeezes the features by dimensions and calculates the motion distribution of each dimension. CAM maintains the predicted texture information, which integrates latent information between channels and weights of the effective channels. Through the coordination of multiple modes, the model has long-term predictive capability and the accuracy of the prediction results has been improved. Experimental results show that using a smaller number of parameters, the proposed method can obtain competitive results but less inference time compared with the state-of-the-art (SOTA) methods. [ABSTRACT FROM AUTHOR]

Automatic modulation recognition is a critical challenge in the field of cognitive radio. In the process of communication, radio signals are modulated in various modes and are interfered by the complex electromagnetic environment. To cope with these problems and avoid manual selection of complex expert features, we propose a multi-level feature extraction algorithm based on deep learning to adequately exploit the hidden feature information of modulated signals. Our algorithm integrates the correlation between the channels of radio signals with convolutional neural networks and Bidirectional Long Short-term Memory (Bi-LSTM), and adopts the appropriate skip connection, which avoids the loss of valid information and achieves the complementarity between spatial and temporal features. In our model, the one-dimensional convolutional layer is specially utilized to enrich the feature representation of each sample point of in-phase and quadrature (I/Q) signals and emphasize the mutual influence of I channel (in-phase signal) and Q channel (quadrature signal). In addition, the label smoothing technique is used to improve the generalization ability of the model. Our proposed method is also of certain significance for other signal processing methods based on deep learning. Experiment results demonstrate that our algorithm outperforms the popular algorithms and is of higher robustness. Specifically, the proposed method improves the recognition accuracy, reaching 92.68% at high signal-to-noise ratio (SNR). In particular, it also reduces the difficulty of recognition for multiple quadrature amplitude modulation (MQAM) signals and significantly improves the recognition accuracy for 16QAM and 64QAM. [ABSTRACT FROM AUTHOR]

MVI is a risk assessment factor related to hepatocellular carcinoma (HCC) recurrence after hepatectomy or liver transplantation. The goal of this paper is to study the preoperative diagnosis of microvascular invasion (MVI) by using a deep learning algorithm in non-contrast T2 weighted magnetic resonance imaging (MRI) images instead of pathological images. Herein, an ensemble learning algorithm named H-DARnet—based on the difference degree and attention mechanism, combined with radiomics, for MVI prediction—is proposed. Our hybrid network combines the fine-grained, high-level semantic, and radiomics features and exhibits a rich multilevel-feature architecture composed of global-local-prior knowledge with suitable complementarity. The total loss function comprises two regularization items––the triplet and the cross-entropy loss function––which are selected for the triplet network and SE-DenseNet, respectively. The hard triplet sample selection strategy for a triplet network and data augmentation for small-scale liver image datasets in convolutional neural network (CNN) training is indispensable. For 200 patch level test samples (135 positive samples and 65 negative samples), our method can obtain the best prediction results, the AUC, sensitivity, and specificity were 0.826, 79.5% and 73.8%, respectively. The experiment results show that MVI can be predicted by using MRI images, and the proposed method is better than other deep learning algorithms and hand-crafted feature algorithms. The proposed ensemble learning algorithm is proved to be an effective method for MVI prediction. • The goal of this paper is to study the preoperative diagnosis of HCC MVI by using a deep learning algorithm on T2WI. • An algorithm named H-DARnet based on the difference degree and attention mechanism, combined with radiomics is proposed. • Our hybrid network exhibits a rich multilevel-feature architecture composed of global-local-prior knowledge. • Hard triplet sample selection strategy and data augmentation for small-scale datasets in CNN training is indispensable. • For 200 patch level test samples, our method can obtain the AUC, sensitivity, and specificity 0.826, 79.5% and 73.8%. [ABSTRACT FROM AUTHOR]