3D object perception in maritime is a promising means to enhance the ability of the autonomous navigation. However, 3D automatic reconstruction of dynamic targets remains a challenging task in the 3D object perception systems. In this paper, we present an innovative GAN-based deep learning neural network for 3D ship reconstruction. The method is able to reconstruct 3D ship models from single view images or videos using shape prior knowledge and a vessel generative adversarial network (Vessel-GAN), which can effectively extract the ship shape features from the single-view images or videos. Furthermore, a 3D feature network optimization model is proposed for loop optimization of 3D ship shape perception. In order to validate the effectiveness of the proposed method, a high resolution single ship dataset (HSSD) was created, and the performance of the proposed method was evaluated against the existing methods, The experimental results show that the Vessel Segment Network (VSNet) can improve the accuracy of the HSSD dataset by 2.5% in the single ship segmentation tasks. In addition, Vessel-GAN was demonstrated to be able to generate multi-view ship images with an impressive FID value of 5.82. Moreover, our model can improve the accuracy of 3D ship modeling by 8.9% on the ShapeNet dataset. This innovative approach promises to enable 3D perception of the marine environment for autonomous ship.

Three-dimensional (3D) point cloud has been employed in a wide range of applications recently. As a powerful weapon for point cloud analysis, point-based point cloud neural networks (PNNs) have demonstrated superior performance with less computation complexity and parameters, compared to sparse 3D convolution-based networks and graph-based convolutional neural networks. However, point-based PNNs still suffer from high computational redundancy, large off-chip memory access, and low parallelism in hardware implementation, thereby hindering the applications on edge devices. In this paper, to address these challenges, an energy-efficient 3D point cloud neural network accelerator is proposed for on-chip edge computing. An efficient filter pruning scheme is used to skip the redundant convolution of pruned filters and zero-value feature channels. A block-wise multi-layer perceptron (MLP) fusion method is proposed to increase the on-chip reuse of features, thereby reducing off-chip memory access. A dual-stream blocking technique is proposed for higher parallelism while maintaining inference accuracy. Implemented in an industrial 28-nm CMOS technology, the proposed accelerator achieves an effective energy efficiency of 12.65 TOPS/W and 0.13 mJ/frame energy consumption for PointNeXt-S at 100 MHz, 0.9 V supply voltage, and 8-bit data width. Compared to the state-of-the-art point cloud neural network accelerators, the proposed accelerator enhances the energy efficiency by up to 66.6× and reduces the energy consumption per frame by up to 70.2×.

Three-dimensional (3-D) understanding or inference has received increasing attention, where 3-D convolutional neural networks (3D-CNNs) have demonstrated superior performance compared to 2D-CNNs, since 3D-CNNs learn features from all three dimensions. However, 3D-CNNs suffer from intensive computation and data movement. In this article, Sagitta, an energy-efficient low-latency on-chip 3D-CNN accelerator, is proposed for edge devices. Locality and small differential value dropout are leveraged to increase the sparsity of activations. A full-zero-skipping convolutional microarchitecture is proposed to fully utilize the sparsity of weights and activations. A hierarchical load-balancing scheme is also introduced to increase the hardware utilization. Specialized architecture and computation flow are proposed to enhance the effectiveness of the proposed techniques. Fabricated in a 55-nm CMOS technology, Sagitta achieves 3.8 TOPS/W for C3D at a latency of 0.1 s and 4.5 TOPS/W for 3D U-Net at a latency of 0.9 s at 100 MHz and 0.91-V supply voltage. Compared to the state-of-the-art 3D-CNN and 2D-CNN accelerators, Sagitta enhances the energy efficiency by up to$379.6\times $and$11\times $, respectively.

An energy-efficient convolutional neural network (CNN) accelerator is proposed for low-power inference on edge devices. An adaptive zero skipping technique is proposed to dynamically skip the zeros in either activations or weights, depending on which has the higher sparsity. The characteristic of non-zero data aggregation is explored to enhance the effectiveness of adaptive zero skipping in performance boosting. To mitigate the load imbalance issue after zero skipping, a sparsity-driven data flow and low-complexity dynamic task allocation are employed for different convolution layers. Facilitated further by a two-stage distiller, the proposed accelerator achieves$5.42\times $,$3.41\times $, and$3.42\times $performance boosting for VGG16, AlexNet, and Mobilenet-v1, respectively, compared to the baseline. Implemented in a 55-nm low power CMOS technology, the proposed accelerator achieves an effective energy efficiency of 2.41 TOPS/W, 2.35 TOPS/W, and 0.64 TOPS/W for VGG16, AlexNet, and Mobilenet-v1, respectively, at 100 MHz and 1.08 V supply voltage.

The monitoring of epilepsy patients in non-hospital environment is highly desirable, where ultra-low power wearable seizure detection devices are essential in such a system. The state-of-the-art epileptic seizure detection algorithms targeting such devices either rely on manual feature extractions, which can be biased due to the experience of experts, or deep neural networks, which suffer from high computation complexity. In this paper, we propose a lightweight deep learning model, LightSeizureNet (LSN), for real-time epileptic seizure detection based on raw EEG data in ultra-low power wearable seizure detection devices. The proposed LSN model includes a patient-independent version and a patient-specific version, both of which avoids manual feature extractions and high computation complexity, while maintaining good classification accuracy. Dilated one-dimensional (1D) convolution, global average pooling, and kernel-wise pruning are adopted to compress the LSN model. The proposed models are evaluated on the CHB-MIT scalp EEG database. The patient-independent LSN model achieves 97.09% accuracy with 6.2M MACs, while the patient-specific LSN model achieves 99.77% accuracy with 3.7 M MACs, which are competitive compared to the state of the art in terms of accuracy and complexity. Furthermore, the proposed model is highly interpretable, which is missing in many previous works. By using a uniform approach to explore the interpretability of the proposed model, fine-grained information such as the activated brain region and the frequency of brainwave during seizures is obtained for clinical diagnosis.

Three-dimensional convolutional neural network (3D-CNN) has demonstrated outstanding classification performance in video recognition compared to two-dimensional CNN (2D-CNN), since 3D-CNN not only learns the spatial features of each frame, but also learns the temporal features across all frames. However, 3D-CNN suffers from intensive computation and data movement. To solve these issues, an energy-efficient low-latency 3D-CNN accelerator is proposed. Temporal locality and small differential value dropout are used to increase the sparsity of activation. Furthermore, to fully utilize the sparsity of weight and activation, a full zero-skipping convolutional microarchitecture is proposed. A hierarchical load-balancing scheme is also introduced to improve resource utilization. With the proposed techniques, a 3D-CNN accelerator is designed in a 55-nm low-power CMOS technology, bringing in up to 9.89× speedup compared to the baseline implementation. Benchmarked with C3D, the proposed accelerator achieves an energy efficiency of 4.66 TOPS/W at 100 MHz and 1.08 V supply voltage.

Utterance classification is a key component in many conversational systems. However, classifying real-world user utterances is challenging, as people may express their ideas and thoughts in manifold ways, and the amount of training data for some categories may be fairly limited, resulting in imbalanced data distributions. To alleviate these issues, we conduct a comprehensive survey regarding data augmentation approaches for text classification, including simple random resampling, word-level transformations, and neural text generation to cope with imbalanced data. Our experiments focus on multi-class datasets with a large number of data samples, which has not been systematically studied in previous work. The results show that the effectiveness of different data augmentation schemes depends on the nature of the dataset under consideration.

Most of the previous 3 DOF (Degree Of Freedom) 3D touch positioning techniques require more than one finger (usually two hands) to be performed, which limits their using space on small mobile devices such as phones and tablets that need one hand to be held in most occasions. Given that the pressure sensitive touch screen would become ubiquitous in near future, we present the pressure-based 3DOF 3D objects positioning and manipulating techniques that only use one hand in operating.