Empowering an intelligent agent with the ability of autonomous navigation in complex and dynamic environments is an important and active research topic in embodied artificial intelligence. In this letter, we address this challenging task from the view of exploiting both the spatial and temporal states of a mobile robot interacting with the crowded environment. Specifically, we propose a Spatial-Temporal State Transformer (ST $^{2}$ ) to encode the states while leveraging the deep reinforcement learning method to find the optimal navigation policy accordingly. Technically, the proposed ST $^{2}$ model consists of a global spatial state encoder and a temporal state encoder, which are built upon the Transformer structure. The spatial state encoder is devised to extract the global spatial features and capture the spatial interaction between pedestrians and the robot. The temporal state encoder is designed to model the temporal correlation among consecutive frames and infer the dynamic relationship of the spatial position transformation. Based on the comprehensive spatial-temporal state representation, the value-based reinforcement learning method is leveraged to obtain the optimal navigation policy. Extensive experiments demonstrate the superiority of the proposed ST $^{2}$ over representative state-of-the-art methods. The source code will be made publicly available.

Depthwise separable convolution (DSC) significantly reduces parameter and floating operations with an acceptable loss of accuracy and has been widely used in various lightweight convolutional neural network (CNN) models. In practical applications, however, DSC accelerators based on graphics processing units (GPUs) cannot fully exploit the performance of DSC and are unsuitable for mobile application scenarios. Moreover, low resource utilization due to idle engines is a common problem in DSC accelerator design. In this paper, a high-performance DSC hardware accelerator based on field-programmable gate arrays (FPGAs) is proposed. A highly reusable and scalable multiplication and accumulation engine is proposed to improve the utilization of computational resources. An efficient convolution algorithm is proposed for depthwise convolution (DWC) and pointwise convolution (PWC), respectively, to reduce the on-chip memory occupancy. Meanwhile, the proposed convolution algorithms achieve partial fusion between PWC and DWC, and improve the off-chip memory access efficiency. To maximise bandwidth utilization and reduce latency when reading feature maps, an address mapping method for off-chip accesses is proposed. The performance of the proposed accelerator is demonstrated by implementing MobileNetV2 on an Intel Arria 10 GX660 FPGA by using Verilog HDL. The experimental results show that the proposed DSC accelerator achieves a performance of 205.1 FPS, 128.8 GFLOPS, and 0.24 GOPS/DSP for input images of size $224\times 224\times 3$.

Agricultural pest and disease monitoring has recently become a crucial aspect of modern agriculture. Toward this end, this study investigates methodologies for implementing low-power, high-performance convolutional neural networks (CNNs) on agricultural edge detection devices. Recognizing the potential of field-programmable gate arrays (FPGAs) to enhance inference parallelism, we leveraged their computational capabilities and intensive storage to propose an embedded FPGA-based CNN accelerator design aimed at optimizing rice leaf disease image classification. Additionally, we trained the MobileNetV2 network using multimodal image data and employed knowledge distillation from a stronger teacher (DIST) as the hardware benchmark. The solution was deployed on the ZYNQ-AC7Z020 hardware platform using High-Level Synthesis (HLS) design tools. Through a combination of fine-grained pipelining, matrix blocking, and linear buffering optimizations, the proposed system achieved a power consumption of 3.21 W, an accuracy of 97.41%, and an inference speed of 43 ms per frame, making it a practical solution for edge-based rice leaf disease classification.

With the attention paid to privacy protection, human motion recognition (HMR) based on radars has attracted more and more attention. A soft synchronization frequency modulated continuous wave (SS-FMCW) radar has a wider application prospect in HMR by virtue of its signal diversity and structural flexibility. In this paper, a new spatiotemporal weighted micro-Doppler (SWMD) spectrum is proposed, which is more competitive for the application of the SS-FMCW radar in HMR. First, to facilitate the understanding of the SWMD spectrum construction method, we introduce the principle behind the SS-FMCW radar based on intermittent ramp signals. Specifically, the SWMD spectrum is constructed by a weighted data fusion algorithm and a time series data fusion algorithm. In addition, we collect data of six different motions from eight testers, and demonstrate and analyze the validity of the SWMD spectrum by multiple dimensions. The experimental results show that compared with the conventional time micro-Doppler spectrum, the accuracy of the SWMD spectrum in HMR is improved by about 7.2%, which is an average of multiple comparison experiments of different classification algorithms with different features.