Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles
The advancement of unmanned aerial vehicles (UAVs) enables early detection of numerous disasters. Efforts have been made to automate the monitoring of data from UAVs, with machine learning methods recently attracting significant interest. These solutions often face challenges with high computational...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-04-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/24/7/2265 |
| _version_ | 1797211956689502208 |
|---|---|
| author | Yu-Pei Liang Chen-Ming Chang Ching-Che Chung |
| author_facet | Yu-Pei Liang Chen-Ming Chang Ching-Che Chung |
| author_sort | Yu-Pei Liang |
| collection | DOAJ |
| description | The advancement of unmanned aerial vehicles (UAVs) enables early detection of numerous disasters. Efforts have been made to automate the monitoring of data from UAVs, with machine learning methods recently attracting significant interest. These solutions often face challenges with high computational costs and energy usage. Conventionally, data from UAVs are processed using cloud computing, where they are sent to the cloud for analysis. However, this method might not meet the real-time needs of disaster relief scenarios. In contrast, edge computing provides real-time processing at the site but still struggles with computational and energy efficiency issues. To overcome these obstacles and enhance resource utilization, this paper presents a convolutional neural network (CNN) model with an early exit mechanism designed for fire detection in UAVs. This model is implemented using TSMC 40 nm CMOS technology, which aids in hardware acceleration. Notably, the neural network has a modest parameter count of 11.2 k. In the hardware computation part, the CNN circuit completes fire detection in approximately 230,000 cycles. Power-gating techniques are also used to turn off inactive memory, contributing to reduced power consumption. The experimental results show that this neural network reaches a maximum accuracy of 81.49% in the hardware implementation stage. After automatic layout and routing, the CNN hardware accelerator can operate at 300 MHz, consuming 117 mW of power. |
| first_indexed | 2024-04-24T10:34:44Z |
| format | Article |
| id | doaj.art-8c92140835ce40eda1f8a889b0266cb8 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-24T10:34:44Z |
| publishDate | 2024-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-8c92140835ce40eda1f8a889b0266cb82024-04-12T13:26:37ZengMDPI AGSensors1424-82202024-04-01247226510.3390/s24072265Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial VehiclesYu-Pei Liang0Chen-Ming Chang1Ching-Che Chung2Department of Computer Science and Information Engineering, Advanced Institute of Manufacturing with High-Tech Innovations, National Chung Cheng University, Chia-Yi 621301, TaiwanDepartment of Computer Science and Information Engineering, Advanced Institute of Manufacturing with High-Tech Innovations, National Chung Cheng University, Chia-Yi 621301, TaiwanDepartment of Computer Science and Information Engineering, Advanced Institute of Manufacturing with High-Tech Innovations, National Chung Cheng University, Chia-Yi 621301, TaiwanThe advancement of unmanned aerial vehicles (UAVs) enables early detection of numerous disasters. Efforts have been made to automate the monitoring of data from UAVs, with machine learning methods recently attracting significant interest. These solutions often face challenges with high computational costs and energy usage. Conventionally, data from UAVs are processed using cloud computing, where they are sent to the cloud for analysis. However, this method might not meet the real-time needs of disaster relief scenarios. In contrast, edge computing provides real-time processing at the site but still struggles with computational and energy efficiency issues. To overcome these obstacles and enhance resource utilization, this paper presents a convolutional neural network (CNN) model with an early exit mechanism designed for fire detection in UAVs. This model is implemented using TSMC 40 nm CMOS technology, which aids in hardware acceleration. Notably, the neural network has a modest parameter count of 11.2 k. In the hardware computation part, the CNN circuit completes fire detection in approximately 230,000 cycles. Power-gating techniques are also used to turn off inactive memory, contributing to reduced power consumption. The experimental results show that this neural network reaches a maximum accuracy of 81.49% in the hardware implementation stage. After automatic layout and routing, the CNN hardware accelerator can operate at 300 MHz, consuming 117 mW of power.https://www.mdpi.com/1424-8220/24/7/2265unmanned aerial vehicles (UAVs)fire detectionneural networksquantizationfixed-point arithmeticreal-time systems |
| spellingShingle | Yu-Pei Liang Chen-Ming Chang Ching-Che Chung Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles Sensors unmanned aerial vehicles (UAVs) fire detection neural networks quantization fixed-point arithmetic real-time systems |
| title | Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles |
| title_full | Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles |
| title_fullStr | Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles |
| title_full_unstemmed | Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles |
| title_short | Implementation of Lightweight Convolutional Neural Networks with an Early Exit Mechanism Utilizing 40 nm CMOS Process for Fire Detection in Unmanned Aerial Vehicles |
| title_sort | implementation of lightweight convolutional neural networks with an early exit mechanism utilizing 40 nm cmos process for fire detection in unmanned aerial vehicles |
| topic | unmanned aerial vehicles (UAVs) fire detection neural networks quantization fixed-point arithmetic real-time systems |
| url | https://www.mdpi.com/1424-8220/24/7/2265 |
| work_keys_str_mv | AT yupeiliang implementationoflightweightconvolutionalneuralnetworkswithanearlyexitmechanismutilizing40nmcmosprocessforfiredetectioninunmannedaerialvehicles AT chenmingchang implementationoflightweightconvolutionalneuralnetworkswithanearlyexitmechanismutilizing40nmcmosprocessforfiredetectioninunmannedaerialvehicles AT chingchechung implementationoflightweightconvolutionalneuralnetworkswithanearlyexitmechanismutilizing40nmcmosprocessforfiredetectioninunmannedaerialvehicles |