{"title":"SF-Yolov8n:用于检测牙钉表面缺陷的新型超轻高精度模型","authors":"Xiaoxin Chen;Zhansi Jiang;Yan Piao;Jingcheng Yang;Hongxin Zheng;Hao Yang;Kequan Chen","doi":"10.1109/JSEN.2024.3392674","DOIUrl":null,"url":null,"abstract":"To improve the detection accuracy of surface defects in dental nails in the current medical manufacturing industry, while reducing the model size and making it easy to deploy on resource-limited devices, we propose a novel model structure for detecting surface defects on dental nails, SF-Yolov8n, based on Yolov8n. Its characteristic is that the model is not only lightweight but also has high detection performance. First, to significantly reduce the number of parameters and volume of the model, while improving the detection accuracy of surface defects on small targets, we simplified the model structure by pruning some network feature layers and modules and added an additional small target detection layer. Second, we developed a new lightweight module, C2Fast_CA, to replace some of the C2f modules in the model to reduce parameters. Afterward, to further simplify the model structure and reduce computational complexity, we made exploratory adjustments to the reg_max in the model to find the minimum value that is most suitable for lightweight models, thereby achieving model miniaturization. Finally, we also optimize the loss function to improve the overall performance of the model in handling various difficult samples. The experimental results show that SF-Yolov8n performs better than other mainstream detection models in detecting surface defects of dental nails and achieves the highest technique for order preference by similarity to an ideal solution (TOPSIS) score. In addition, the parameter quantity of SF-Yolov8n is only 0.69 M, which is a reduction of 77.01% compared with Yolov8n. Meanwhile, the including precision (P), recall (R), and mAP50 of SF-Yolov8n are increased by 3.7%, 3.4%, and 5.8%, respectively, compared with Yolov8n.","PeriodicalId":447,"journal":{"name":"IEEE Sensors Journal","volume":null,"pages":null},"PeriodicalIF":4.3000,"publicationDate":"2024-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"SF-Yolov8n: A Novel Ultralightweight and High-Precision Model for Detecting Surface Defects of Dental Nails\",\"authors\":\"Xiaoxin Chen;Zhansi Jiang;Yan Piao;Jingcheng Yang;Hongxin Zheng;Hao Yang;Kequan Chen\",\"doi\":\"10.1109/JSEN.2024.3392674\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"To improve the detection accuracy of surface defects in dental nails in the current medical manufacturing industry, while reducing the model size and making it easy to deploy on resource-limited devices, we propose a novel model structure for detecting surface defects on dental nails, SF-Yolov8n, based on Yolov8n. Its characteristic is that the model is not only lightweight but also has high detection performance. First, to significantly reduce the number of parameters and volume of the model, while improving the detection accuracy of surface defects on small targets, we simplified the model structure by pruning some network feature layers and modules and added an additional small target detection layer. Second, we developed a new lightweight module, C2Fast_CA, to replace some of the C2f modules in the model to reduce parameters. Afterward, to further simplify the model structure and reduce computational complexity, we made exploratory adjustments to the reg_max in the model to find the minimum value that is most suitable for lightweight models, thereby achieving model miniaturization. Finally, we also optimize the loss function to improve the overall performance of the model in handling various difficult samples. The experimental results show that SF-Yolov8n performs better than other mainstream detection models in detecting surface defects of dental nails and achieves the highest technique for order preference by similarity to an ideal solution (TOPSIS) score. In addition, the parameter quantity of SF-Yolov8n is only 0.69 M, which is a reduction of 77.01% compared with Yolov8n. Meanwhile, the including precision (P), recall (R), and mAP50 of SF-Yolov8n are increased by 3.7%, 3.4%, and 5.8%, respectively, compared with Yolov8n.\",\"PeriodicalId\":447,\"journal\":{\"name\":\"IEEE Sensors Journal\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":4.3000,\"publicationDate\":\"2024-04-30\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Sensors Journal\",\"FirstCategoryId\":\"103\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10516300/\",\"RegionNum\":2,\"RegionCategory\":\"综合性期刊\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Sensors Journal","FirstCategoryId":"103","ListUrlMain":"https://ieeexplore.ieee.org/document/10516300/","RegionNum":2,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
SF-Yolov8n: A Novel Ultralightweight and High-Precision Model for Detecting Surface Defects of Dental Nails
To improve the detection accuracy of surface defects in dental nails in the current medical manufacturing industry, while reducing the model size and making it easy to deploy on resource-limited devices, we propose a novel model structure for detecting surface defects on dental nails, SF-Yolov8n, based on Yolov8n. Its characteristic is that the model is not only lightweight but also has high detection performance. First, to significantly reduce the number of parameters and volume of the model, while improving the detection accuracy of surface defects on small targets, we simplified the model structure by pruning some network feature layers and modules and added an additional small target detection layer. Second, we developed a new lightweight module, C2Fast_CA, to replace some of the C2f modules in the model to reduce parameters. Afterward, to further simplify the model structure and reduce computational complexity, we made exploratory adjustments to the reg_max in the model to find the minimum value that is most suitable for lightweight models, thereby achieving model miniaturization. Finally, we also optimize the loss function to improve the overall performance of the model in handling various difficult samples. The experimental results show that SF-Yolov8n performs better than other mainstream detection models in detecting surface defects of dental nails and achieves the highest technique for order preference by similarity to an ideal solution (TOPSIS) score. In addition, the parameter quantity of SF-Yolov8n is only 0.69 M, which is a reduction of 77.01% compared with Yolov8n. Meanwhile, the including precision (P), recall (R), and mAP50 of SF-Yolov8n are increased by 3.7%, 3.4%, and 5.8%, respectively, compared with Yolov8n.
期刊介绍:
The fields of interest of the IEEE Sensors Journal are the theory, design , fabrication, manufacturing and applications of devices for sensing and transducing physical, chemical and biological phenomena, with emphasis on the electronics and physics aspect of sensors and integrated sensors-actuators. IEEE Sensors Journal deals with the following:
-Sensor Phenomenology, Modelling, and Evaluation
-Sensor Materials, Processing, and Fabrication
-Chemical and Gas Sensors
-Microfluidics and Biosensors
-Optical Sensors
-Physical Sensors: Temperature, Mechanical, Magnetic, and others
-Acoustic and Ultrasonic Sensors
-Sensor Packaging
-Sensor Networks
-Sensor Applications
-Sensor Systems: Signals, Processing, and Interfaces
-Actuators and Sensor Power Systems
-Sensor Signal Processing for high precision and stability (amplification, filtering, linearization, modulation/demodulation) and under harsh conditions (EMC, radiation, humidity, temperature); energy consumption/harvesting
-Sensor Data Processing (soft computing with sensor data, e.g., pattern recognition, machine learning, evolutionary computation; sensor data fusion, processing of wave e.g., electromagnetic and acoustic; and non-wave, e.g., chemical, gravity, particle, thermal, radiative and non-radiative sensor data, detection, estimation and classification based on sensor data)
-Sensors in Industrial Practice