Feature Map Data Research Articles

Open-Pit Granite Mining Area Extraction Using UAV Aerial Images and the Novel GIPNet

The ability to rapidly and accurately delineate open-pit granite mining areas is pivotal for effective production planning and environmental impact assessment. Over the years, advancements in remote sensing techniques, including the utilization of satellite imagery, LiDAR technology and unmanned aerial vehicles, have revolutionized the way mining areas are monitored and managed. Simultaneously, in the context of the open-pit mining area extraction task, deep learning-based automatic recognition is gradually replacing manual visual interpretation. Leveraging the potential of unmanned aerial vehicles (UAVs) for real-time, low-risk remote sensing, this study employs UAV-derived orthophotos for mining area extraction. Central to the proposed approach is the novel Gather–Injection–Perception (GIP) module, designed to overcome the information loss typically associated with conventional feature pyramid modules during feature fusion. The GIP module effectively enriches semantic features, addressing a crucial information limitation in existing methodologies. Furthermore, the network introduces the Boundary Perception (BP) module, uniquely tailored to tackle the challenges of blurred boundaries and imprecise localization in mining areas. This module capitalizes on attention mechanisms to accentuate critical high-frequency boundary details in the feature map and synergistically utilizes both high- and low-dimensional feature map data for deep supervised learning. The suggested method demonstrates its superiority in a series of comparative experiments on a specially assembled dataset of research area images. The results are compelling, with the proposed approach achieving 90.67% precision, 92.00% recall, 91.33% F1-score, and 84.04% IoU. These figures not only underscore the effectiveness of suggested model in enhancing the extraction of open-pit granite mining areas but also provides a new idea for the subsequent application of UAV data in the mining scene.

IoT Device Authentication Using Self-Organizing Feature Map Data Sets

Sensors and actuators connected via the Internet of Things (IoT) have now become embedded within our critical infrastructure offering improved observation and control as well as reduced costs. Given that software defined radios (SDRs) can be readily programmed to imitate IoT devices, there is now a greater risk that assets can be spoofed or compromised. This necessitates an urgent need for IoT device authentication, avoiding the need to upgrade the many thousands of individual devices. However, the lack of publicly available data sets severely hampers the development of effective authentication algorithms and mechanisms. In this regard, this article introduces a technique for facilitating IoT device authentication when the radio frequency (RF) characteristics are highly correlated using self-organizing feature maps (SOFMs), thus aiming to promote state-of-the-art research in this field. The associated techniques demonstrated in this article exploit a novel data set of RF fingerprints and are, in particular, suitable for low-cost and long-range wireless application scenarios of the IoT, e.g., LoRa. Here, a well trained convolutional neural network (CNN) based on the SOFM data set can rapidly profile apparently correlated RF fingerprint patterns and thereby ascertain the nature of a specific device (friend or foe). In this way, a reliable and efficient IoT device authentication strategy for LoRa devices can be established. The experimental results presented in this article substantiate the effectiveness and efficiency of the SOFM based approach, and the data sets are introduced in detail and shared with the research community.

Efficient Binary Weight Convolutional Network Accelerator for Speech Recognition.

Speech recognition has progressed tremendously in the area of artificial intelligence (AI). However, the performance of the real-time offline Chinese speech recognition neural network accelerator for edge AI needs to be improved. This paper proposes a configurable convolutional neural network accelerator based on a lightweight speech recognition model, which can dramatically reduce hardware resource consumption while guaranteeing an acceptable error rate. For convolutional layers, the weights are binarized to reduce the number of model parameters and improve computational and storage efficiency. A multichannel shared computation (MCSC) architecture is proposed to maximize the reuse of weight and feature map data. The binary weight-sharing processing engine (PE) is designed to avoid limiting the number of multipliers. A custom instruction set is established according to the variable length of voice input to configure parameters for adapting to different network structures. Finally, the ping-pong storage method is used when the feature map is an input. We implemented this accelerator on Xilinx ZYNQ XC7Z035 under the working frequency of 150 MHz. The processing time for 2.24 s and 8 s of speech was 69.8 ms and 189.51 ms, respectively, and the convolution performance reached 35.66 GOPS/W. Compared with other computing platforms, accelerators perform better in terms of energy efficiency, power consumption and hardware resource consumption.

A Real-Time Convolutional Neural Network for Super-Resolution on FPGA With Applications to 4K UHD 60 fps Video Services

In this paper, we present a novel hardware-friendly super-resolution (SR) method based on a convolutional neural network (CNN) and its dedicated hardware (HW) on field programmable gate array (FPGA). Although CNN-based SR methods have shown very promising results for SR, their computational complexities are prohibitive for hardware implementation. To the best of our knowledge, we are the first to implement a real-time CNN-based SR HW that upscales 2K full high-definition video to 4K ultra high-definition (UHD) video at 60 frames per second (fps). In our dedicated CNN-based SR HW, low-resolution input frames are processed line-by-line, and the number of convolutional filter parameters is reduced significantly by incorporating depth-wise separable convolutions with a residual connection. Our CNN-based SR HW incorporates a cascade of 1D convolutions having large receptive fields along horizontal lines while keeping vertical receptive fields minimal, which allows us to save required line memory space in achieving comparable SR performance against full 2D convolution operations. For efficient HW implementation, we use a simple and effective quantization method with little peak signal-to-noise ratio (PSNR) degradation. Also, we propose a compression method to efficiently store intermediate feature map data to reduce the number of line memories used in HW. Our HW implementation on the FPGA generates 4K UHD frames of higher PSNR values at 60 fps and shows better visual quality, compared with conventional CNN-based SR methods that are trained and tested in software.

Evaluating Fast Algorithms for Convolutional Neural Networks on FPGAs

In recent years, convolutional neural networks (CNNs) have become widely adopted for computer vision tasks. Field-programmable gate arrays (FPGAs) have been adequately explored as a promising hardware accelerator for CNNs due to its high performance, energy efficiency, and reconfigurability. However, prior FPGA solutions based on the conventional convolutional algorithm is often bounded by the computational capability of FPGAs (e.g., the number of DSPs). To address this problem, the feature maps are transformed to a special domain using fast algorithms to reduce the arithmetic complexity. Winograd and fast Fourier transformation (FFT), as fast algorithm representatives, first transform input data and filter to Winograd or frequency domain, then perform element-wise multiplication, and apply inverse transformation to get the final output. In this paper, we propose a novel architecture for implementing fast algorithms on FPGAs. Our design employs line buffer structure to effectively reuse the feature map data among different tiles. We also effectively pipeline the Winograd/FFT processing element (PE) engine and initiate multiple PEs through parallelization. Meanwhile, there exists a complex design space to explore. We propose an analytical model to predict the resource usage and the performance. Then, we use the model to guide a fast design space exploration. Experiments using the state-of-the-art CNNs demonstrate the best performance and energy efficiency on FPGAs. We achieve 854.6 and 2479.6 GOP/s for AlexNet and VGG16 on Xilinx ZCU102 platform using Winograd. We achieve 130.4 GOP/s for Resnet using Winograd and 201.1 GOP/s for YOLO using FFT on Xilinx ZC706 platform.

The Implementation of Chinese Acoustic Model Efficiency Optimization Based on Convolutional Neural Network

In the speech recognition system, the traditional convolution method of convolutional neural network (CNN) model has the problem of low efficiency, especially when processing a large number of speech feature map data. In this paper, a new convolution method for the CNN model has been proposed, which makes the efficiency of CNN model about five times faster than the traditional method. And at the same time, the recognition accuracy of the new method is comparable with traditional method.

Research on Household Appliances Recognition Method Based on Data Screening of Deep Learning

The Non-intrusive household load indentification can realize a series of power quality analysis such as power management, energy monitoring. It has the advantages of low cost, easy implementation. Aiming at the problem that a large number of V-I trajectory sample data is unavailable due to noise interference of stable operation data of household appliances of the actual measurement, a V-I map sample data set screening algorithm is proposed, which screen the two-dimensional V-I feature map data sets of characterizing household appliances, improves the deep learning network, and achieves better recognition effect of household appliances by using transfer learning. Experiments show that this method can effectively improve the accuracy of load identification algorithm, and has more advantages than traditional methods.