Scholarlink

<p>VLSI system design for the Internet of Things (IoT) offers various opportunities beyond conventional semiconductor applications. Big chips are the focus of traditional system-on-chip design, whereas low cost and low power consumption are the focus of IoT device design.VLSI design for IoT requires novel mindset-big chips are not best fit for edge and fog devices. The Progressive Cyclical Convolutional Neural Network for embedded vision based Internet of Things (IoT) using VLSI is proposed (PCCNN-IOT-VLSI) in this manuscript. Initially, the embedded vision for IoT applications are developed using PCCNN. Then, Corona-virus Mask Protection Algorithm (CMPA) is used to optimize the input weight parameters of the PCCNN. The proposed PCCNN-IOT-VLSI is implemented in MATLAB and its performance is analyzed with the help of performance metrics such as, computational complexity, hardware complexity, critical path delay, storage complexity, bandwidth, dissipation, throughput, accuracy. The proposed PCCNN-IOT-VLSI method provides 24.53%, 28.87%, 32.34% higher accuracy, 25.67%, 22.66%, 27.92% lower computational complexity when compared to the existing methods: Investigation into designing VLSI of a flexible architecture for a deep neural network accelerator (DNNA-IOT-VLSI), S2RNN: Self-Supervised Reconfigurable Neural Network Hardware Accelerator for Machine Learning Applications (S2RNN-IOT-VLSI), and Towards reconfigurable CNN accelerator for FPGA implementation (CNN-IOT-VLSI) respectively.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Progressive Cyclical Convolutional Neural Network for embedded vision based Internet of Things using VLSI

  • 期刊论文

  • 发表日期：

    2025年10月22日

查看全文
  • S. Jayakumar,
  • Koduru Gouthami ,
  • K. Chanthirasekaran,
  • Kanthapitchai Paul Joshua

摘要

VLSI system design for the Internet of Things (IoT) offers various opportunities beyond conventional semiconductor applications. Big chips are the focus of traditional system-on-chip design, whereas low cost and low power consumption are the focus of IoT device design.VLSI design for IoT requires novel mindset-big chips are not best fit for edge and fog devices. The Progressive Cyclical Convolutional Neural Network for embedded vision based Internet of Things (IoT) using VLSI is proposed (PCCNN-IOT-VLSI) in this manuscript. Initially, the embedded vision for IoT applications are developed using PCCNN. Then, Corona-virus Mask Protection Algorithm (CMPA) is used to optimize the input weight parameters of the PCCNN. The proposed PCCNN-IOT-VLSI is implemented in MATLAB and its performance is analyzed with the help of performance metrics such as, computational complexity, hardware complexity, critical path delay, storage complexity, bandwidth, dissipation, throughput, accuracy. The proposed PCCNN-IOT-VLSI method provides 24.53%, 28.87%, 32.34% higher accuracy, 25.67%, 22.66%, 27.92% lower computational complexity when compared to the existing methods: Investigation into designing VLSI of a flexible architecture for a deep neural network accelerator (DNNA-IOT-VLSI), S2RNN: Self-Supervised Reconfigurable Neural Network Hardware Accelerator for Machine Learning Applications (S2RNN-IOT-VLSI), and Towards reconfigurable CNN accelerator for FPGA implementation (CNN-IOT-VLSI) respectively.