高性能YOLOv5：面向嵌入式平臺(tái)高性能目標(biāo)檢測(cè)算法研究

劉喬壽; 趙志源; 王均成; 皮勝文

doi:10.11999/JEIT220413

高性能YOLOv5：面向嵌入式平臺(tái)高性能目標(biāo)檢測(cè)算法研究

doi: 10.11999/JEIT220413

1.
重慶郵電大學(xué)通信與信息工程學(xué)院重慶 400065
2.
先進(jìn)網(wǎng)絡(luò)與智能互聯(lián)技術(shù)重慶市高校重點(diǎn)實(shí)驗(yàn)室重慶 400065
3.
泛在感知與互聯(lián)重慶市重點(diǎn)實(shí)驗(yàn)室重慶 400065

詳細(xì)信息

作者簡(jiǎn)介:
劉喬壽：男，副教授，博士，研究方向?yàn)?G超密集網(wǎng)絡(luò)干擾協(xié)調(diào)、云邊協(xié)同智能計(jì)算、FPGA智能算法加速、物聯(lián)網(wǎng)系統(tǒng)及終端設(shè)備開發(fā)

趙志源：男，碩士生，研究方向?yàn)槟繕?biāo)檢測(cè)和云邊協(xié)同智能計(jì)算

王均成：男，碩士生，研究方向?yàn)槟繕?biāo)檢測(cè)和圖像處理

皮勝文：男，碩士生，研究方向?yàn)榉植际缴疃葘W(xué)習(xí)與通算融合

通訊作者:
趙志源　804311347@qq.com

中圖分類號(hào): TN911.73
計(jì)量
- 文章訪問數(shù): 2136
- HTML全文瀏覽量: 2019
- PDF下載量: 445
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2022-04-07
- 修回日期: 2022-07-07
- 網(wǎng)絡(luò)出版日期: 2022-07-08
- 刊出日期: 2023-06-10

High Performance YOLOv5: Research on High Performance Target Detection Algorithm for Embedded Platform

1.
School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
2.
Advanced Network and Intelligent Connection Technology Key Laboratory of Chongqing Education Commission of China, Chongqing 400065, China
3.
Chongqing Key Laboratory of Ubiquitous Sensing and Networking, Chongqing 400065, China

摘要

摘要: 針對(duì)目前深度學(xué)習(xí)單階段檢測(cè)算法綜合性能不平衡以及在嵌入式設(shè)備難以部署等問題，該文提出一種面向嵌入式平臺(tái)的高性能目標(biāo)檢測(cè)算法?；谥豢?次5代 (YOLOv5)網(wǎng)絡(luò)，改進(jìn)算法首先在主干網(wǎng)絡(luò)部分采用設(shè)計(jì)的空間頸塊代替原有的焦點(diǎn)模塊，結(jié)合改進(jìn)的混洗網(wǎng)絡(luò)2代替換原有的跨級(jí)局部暗網(wǎng)絡(luò)，減小空間金字塔池化 (SPP)的內(nèi)核尺寸，實(shí)現(xiàn)了主干網(wǎng)絡(luò)的輕量化。其次，頸部采用了基于路徑聚合網(wǎng)絡(luò) (PAN)設(shè)計(jì)的增強(qiáng)型路徑聚合網(wǎng)絡(luò) (EPAN)，增加了P6大目標(biāo)輸出層，提高了網(wǎng)絡(luò)的特征提取能力。然后，檢測(cè)頭部分采用以自適應(yīng)空間特征融合 (ASFF)為基礎(chǔ)設(shè)計(jì)的自適應(yīng)空洞空間特征融合 (A-ASFF)來替代原有的檢測(cè)頭，解決了物體尺度變化問題，在少量增加額外開銷情況下大幅提升檢測(cè)精度。最后，函數(shù)部分采用高效交并比 (EIoU)代替完整交并比 (CIoU)損失函數(shù)，采用S型加權(quán)線性單元 (SiLU)代替HardSwish激活函數(shù)，提升了模型的綜合性能。實(shí)驗(yàn)結(jié)果表明，與YOLOv5-S相比，該文提出的同版本算法在mAP@.5，mAP@.5:.95上分別提高了4.6%和6.3%，參數(shù)量降低了43.5%，計(jì)算復(fù)雜度降低了12.0%，在Jetson Nano平臺(tái)上使用原模型和TensorRT加速模型進(jìn)行速度評(píng)估，分別減少了8.1%和9.8%的推理延遲。該文所提算法的綜合指標(biāo)超越了眾多優(yōu)秀的目標(biāo)檢測(cè)網(wǎng)絡(luò)，對(duì)嵌入式平臺(tái)更為友好，具有實(shí)際應(yīng)用意義。
- 目標(biāo)檢測(cè) /
- YOLOv5 /
- 混洗網(wǎng)絡(luò)2代 /
- 自適應(yīng)空間特征融合 /
- 嵌入式設(shè)備 /
- TensorRT加速
Abstract: Considering the problems of imbalanced comprehensive performance of the current deep learning single-stage detection algorithms and difficult deployment in embedded devices, one High-Performance object detection algorithm for embedded platforms is proposed in this paper. Based on the You Only Look Once v5 (YOLOv5) network, in the backbone network part of the improved algorithm firstly, the original focus module and original Cross Stage Partial Darknet are replaced by a designed space stem block and an improved ShuffleNetv2, respectively. The kernel size of Space Pyramid Pooling (SPP) is reduced to lighten the backbone network. Secondly, in the neck, an Enhanced Path Aggregation Network (EPAN) based on Path Aggregation Network (PAN) design is adopted, a P6 large target output layer is added, and the feature extraction ability of the network is improved. And then, in the head, an Adaptive-Atrous Spatial Feature Fusion (A-ASFF) based on Adaptive Spatial Feature Fusion (ASFF) is used to replace the original detection head, the object scale change problem is solved, and the detection accuracy is greatly improved with a small amount of additional overhead. Finally, in the function section, a Complete Intersection over Union (CIoU) loss function is replaced by the Efficient Intersection over Union (EIoU), a HardSwish activation function is replaced by a Sigmoid weighted Linear Unit (SiLU), and model synthesis ability has been improved. The experimental results show that compared to YOLOv5-S, the mAP@.5 and mAP@.5:95 of the same version of the algorithm proposed in this paper are increased by 4.6% and 6.3% while the number of parameters and the computational complexity are reduced by 43.5% and 12.0%, respectively. Using the original model and the TensorRT accelerated model for speed evaluation on the Jetson Nano platform, the inference latency is reduced by 8.1% and 9.8%, respectively. The comprehensive indicators of many excellent object detection networks and their friendliness to embedded platforms are surpassed by the algorithm proposed in this paper and the practical meaning is generated.
- Object detection /
- YOLOv5 /
- ShuffleNetv2 /
- Adaptive Spatial Feature Fusion (ASFF) /
- Embedded device /
- TensorRT acceleration

HTML全文

圖 1 YOLOv5整體架構(gòu)圖

下載: 全尺寸圖片幻燈片

圖 2 Space Stem架構(gòu)圖

下載: 全尺寸圖片幻燈片

圖 3 Shuffle Block和IS Block的模塊對(duì)比圖

下載: 全尺寸圖片幻燈片

圖 4 PAN與EPAN的模塊對(duì)比圖

下載: 全尺寸圖片幻燈片

圖 6 A-ASFF Head結(jié)構(gòu)

下載: 全尺寸圖片幻燈片

圖 5 HP-YOLOv5整體架構(gòu)圖

下載: 全尺寸圖片幻燈片

圖 7 YOLOv5-S和 HP-YOLOv5-S的平均精度和召回率對(duì)比圖

下載: 全尺寸圖片幻燈片

圖 8 YOLOv5-S不同場(chǎng)景檢測(cè)效果圖

下載: 全尺寸圖片幻燈片

圖 9 HP-YOLOv5-S不同場(chǎng)景檢測(cè)效果圖

下載: 全尺寸圖片幻燈片

表 1 網(wǎng)絡(luò)疊加優(yōu)化研究結(jié)果

方法	mAP@.5(%)	mAP@.5:.95(%)	Params(M)	FLOPs(G)	Latency (ms)
YOLOv5	69.7	42.1	7.3	8.3	11.2
替換SiLU	70.2(+0.5)	42.4(+0.3)	7.3	8.3	10.8(–0.4)
修改SPP(3,5,9)	70.2	42.4	7.3	8.3	10.3(–0.5)
替換Space Stem	71.5(+1.3)	44.1(+1.7)	7.3	8.4	10.5(+0.2)
替換IS Block	70.8(–0.7)	43.2(–0.9)	3.8	7.0	8.3(–2.2)
替換EPAN	71.8(+1.0)	44.0(+0.8)	3.9	7.1	8.7(+0.4)
增添P6 Block	73.0(+1.2)	45.5(+1.5)	4.0	7.1	8.9(+0.2)
替換A-ASFF/ASFF	74.1/74.3(+1.1/+1.3)	47.6/47.7(+2.1/+2.2)	4.1/6.6	7.3/10.1	10.1/11.7(+1.2/2.8)
替換EIoU	74.3(+0.2)	48.4(+0.8)	4.1	7.3	10.1

下載: 導(dǎo)出CSV

表 2 本文方法與基線各模塊的對(duì)比實(shí)驗(yàn)結(jié)果

原始方法/本文方法	Params(修改前/后)	MFLOPs(修改前/后)	mAP@.5(修改前/后)(%)
SPP(5,9,13) / SPP(3,5,9)	656896.0/656896.0	257.8/257.8	69.7/69.7
Focus / Space Stem	44032.0/67968.0(+54.0%)	1643.0/2497.0(+51.9%)	69.7/71.3(+1.6)
CSP Block / IS Block	986112.0/882432.0(–10.5%)	1549.0/71.3(–95.4%)	69.7/69.4(–0.3)
PAN / EPAN	193654200.0/200566200.0(+3.6%)	5838.6/5841.4(+0.1%)	69.7/70.8(+1.1)
Head / A-ASFF Head	158300900.0/175772300.0(+11.0%)	1442.7/1719.3(+19.2%)	69.7/71.6(+1.9)

下載: 導(dǎo)出CSV

表 3 本文方法與基線消融對(duì)比實(shí)驗(yàn)結(jié)果

SiLU	Space Stem	IS Block	SPP (3,5,9)	EPAN	P6	A-ASFF	EIOU	mAP@.5(%)	mAP@.5:.95(%)	Params(M)	FLOPs(G)	Latency (ms)
								69.7	42.1	7.3	8.3	11.2
√								70.2↑	42.4↑	7.3	8.3	10.8↑
	√							71.3↑	43.9↑	7.3	8.4↓	11.5↓
		√						69.4↓	41.3↓	3.8↑	6.9↑	8.9↑
			√					69.7	42.1	7.3	8.3	10.7↑
				√				70.8↑	43.3↑	7.4↓	8.4↓	11.6↓
					√			71.6↑	44.9↑	7.5↓	8.4↓	11.4↓
						√		71.6↑	45.3↑	7.6↓	8.7↓	12.5↓
							√	70.0↑	42.9↑	7.3	8.3	11.2
	√		√					71.3↑	43.9↑	7.3	8.4↓	11.1↑
		√	√					69.4↓	41.3↓	3.8↑	6.9↑	8.4↑
	√	√						70.8↑	43.2↑	3.8↑	7.0↑	9.1↑
	√	√	√					70.8↑	43.2↑	3.8↑	7.0↑	8.6↑
				√	√			72.5↑	45.7↑	7.6↓	8.4↓	11.7↓
√	√	√	√	√	√	√	√	74.3↑	48.4↑	4.1↑	7.3↑	10.1↑

下載: 導(dǎo)出CSV

表 4 在PASCAL VOC 2012數(shù)據(jù)集上不同算法性能比較

算法	Params(M)	FLOPs(G)	mAP@.5(%)	mAP@.5:.95(%)	Latency/(ms)			模型大小 (MB)
算法	Params(M)	FLOPs(G)	mAP@.5(%)	mAP@.5:.95(%)	PC	Nano	TRT	模型大小 (MB)
YOLOv3-Tiny	8.7	6.4	47.9	21.5	–	–	–	33.1
YOLOv4-Tiny	5.9	8.0	54.8	25.6	–	–	–	23.7
NanoDet	0.9	1.4	54.4	25.2	8.5	–	–	1.7
NanoDet-Plus	1.2	1.8	57.1	29.3	9.0	–	–	2.3
YOLOv5-N	1.9	2.2	61.9	34.2	9.2	115.2	25.8	4.0
YOLOv5-S	7.3	8.3	69.7	42.1	11.3	152.8	45.7	14.9
YOLOv5-M	21.6	25.3	75.1	49.8	13.2	315.6	97.4	43.5
HP-YOLOv5-N	1.1	1.9	67.2	39.6	8.9	108.4	23.9	3.1
HP-YOLOv5-S	4.1	7.3	74.3	48.4	10.1	140.4	41.2	11.4
HP-YOLOv5-M	11.0	19.1	76.7	52.7	12.5	284.5	86.1	35.6

下載: 導(dǎo)出CSV

參考文獻(xiàn)(20)

[1]	羅會(huì)蘭, 陳鴻坤. 基于深度學(xué)習(xí)的目標(biāo)檢測(cè)研究綜述[J]. 電子學(xué)報(bào), 2020, 48(6): 1230–1239. doi: 10.3969/j.issn.0372-2112.2020.06.026 LUO Huilan and CHEN Hongkun. Survey of object detection based on deep learning[J]. Acta Electronica Sinica, 2020, 48(6): 1230–1239. doi: 10.3969/j.issn.0372-2112.2020.06.026
[2]	GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]. IEEE Conference on Computer Vision and Pattern Recognition, Columbus, USA, 2014: 580–587.
[3]	LIU Wei, ANGUELOV D, ERHAN D, et al. SSD: Single shot MultiBox detector[C]. The 14th European Conference on Computer Vision, Amsterdam, The Netherlands, 2016: 21–37.
[4]	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real-time object detection[C]. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016: 779–788.
[5]	REDMON J and FARHADI A. YOLO9000: Better, faster, stronger[C]. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, USA, 2017: 6517–6525.
[6]	REDMON J and FARHADI A. YOLOv3: An incremental improvement[EB/OL]. http://arxiv.org/abs/1804.02767, 2018.
[7]	BOCHKOVSKIY A, WANG C Y, and LIAO H Y M. YOLOv4: Optimal speed and accuracy of object detection[EB/OL]. https://arxiv.org/abs/2004.10934v1, 2020.
[8]	ULTRALYTICS. YOLOv5[EB/OL]. https://github.com/ultralytics/YOLOv5, 2021.
[9]	賴潤平, 周鵬程, 張梓赫, 等. 基于Jetson Nano的目標(biāo)跟蹤小車的設(shè)計(jì)與實(shí)現(xiàn)[J]. 現(xiàn)代信息科技, 2021, 5(4): 183–187. doi: 10.19850/j.cnki.2096-4706.2021.04.046 LAI Runping, ZHOU Pengcheng, ZHANG Zihe, et al. Design and implementation of object tracking car based on Jetson Nano[J]. Modern Information Technology, 2021, 5(4): 183–187. doi: 10.19850/j.cnki.2096-4706.2021.04.046
[10]	王文勝, 李繼旺, 吳波, 等. 基于YOLOv5交通標(biāo)志識(shí)別的智能車設(shè)計(jì)[J]. 國外電子測(cè)量技術(shù), 2021, 40(10): 158–164. doi: 10.19652/j.cnki.femt.2102913 WANG Wensheng, LI Jiwang, WU Bo, et al. Smart car design based on traffic sign recognition via YOLOv5[J]. Foreign Electronic Measurement Technology, 2021, 40(10): 158–164. doi: 10.19652/j.cnki.femt.2102913
[11]	WANG Xiangheng, YUE Xuebin, LI Hengyi, et al. A high-efficiency dirty-egg detection system based on YOLOv4 and TensorRT[C]. 2021 International Conference on Advanced Mechatronic Systems (ICAMechS), Tokyo, Japan, 2021: 75–80.
[12]	WANG C Y, BOCHKOVSKIY A, and LIAO H Y M. Scaled-yolov4: Scaling cross stage partial network[C]. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nashville, USA, 2021: 13024–13033.
[13]	MA Ningning, ZHANG Xiangyu, ZHENG Haitao, et al. Shufflenet V2: Practical guidelines for efficient CNN architecture design[C]. The 15th European Conference on Computer Vision, Munich, Germany, 2018: 116–131.
[14]	RANGILYU. NanoDet[EB/OL]. https://github.com/RangiLyu/nanodet, 2021.
[15]	LIU Songtao, HUANG Dihuang, and WANG Yunhong. Learning spatial fusion for single-shot object detection[EB/OL]. https://arxiv.org/abs/1911.09516, 2019.
[16]	WANG R J, LI Xiang, and LING C X. Pelee: A real-time object detection system on mobile devices[EB/OL]. https://arxiv.org/abs/1804.06882, 2019.
[17]	HAN Kai, WANG Yunhe, TIAN Qi, et al. GhostNet: More features from cheap operations[C]. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2020: 1577–1586.
[18]	QI Delong, TAN Weijun, YAO Qi, et al. YOLO5Face: Why reinventing a face detector[EB/OL]. https://arxiv.org/abs/2105.12931, 2022.
[19]	ZHANG Yifan, REN Weiqiang, ZHANG Zhang, et al. Focal and efficient IOU loss for accurate bounding box regression[EB/OL]. https://arxiv.org/abs/2101.08158, 2021.
[20]	ELFWING S, UCHIBE E, and DOYA K. Sigmoid-weighted linear units for neural network function approximation in reinforcement learning[J]. Neural Networks, 2018, 107: 3–11. doi: 10.1016/j.neunet.2017.12.012