Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774491
Quoc-Thang Phan, Yuan-Kang Wu, Q. Phan
Solar power forecasting has already become a key role in energy market. However, forecasting PV generation is a challenging task because solar energy strongly depends on weather conditions, day/night cycles and meteorological variables such as solar irradiance, temperature, humidity, etc. Therefore, integrating solar power into grids requires accurate prediction methods, which is also improve the quality of the power systems and reduce operational cost. Beside traditional approaches for forecasting solar PV power generation, new techniques arise every year to enhance the performance of models with the main objective to reduce uncertainties. In this paper, a novel deep learning model based on transformer has been established for one-hour-ahead PV generation forecasting. In practice, this work uses PV power output data with one-hour resolution from the North PV sites in Taiwan, and 2 years of Numerical Weather Prediction (NWP) data from Central Weather Bureau (CWB). Furthermore, forecasting results show that the proposed model outperforms other models such as Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) in terms of NRMSE and NMAPE.
{"title":"An Approach Using Transformer-based Model for Short-term PV generation forecasting","authors":"Quoc-Thang Phan, Yuan-Kang Wu, Q. Phan","doi":"10.1109/ICASI55125.2022.9774491","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774491","url":null,"abstract":"Solar power forecasting has already become a key role in energy market. However, forecasting PV generation is a challenging task because solar energy strongly depends on weather conditions, day/night cycles and meteorological variables such as solar irradiance, temperature, humidity, etc. Therefore, integrating solar power into grids requires accurate prediction methods, which is also improve the quality of the power systems and reduce operational cost. Beside traditional approaches for forecasting solar PV power generation, new techniques arise every year to enhance the performance of models with the main objective to reduce uncertainties. In this paper, a novel deep learning model based on transformer has been established for one-hour-ahead PV generation forecasting. In practice, this work uses PV power output data with one-hour resolution from the North PV sites in Taiwan, and 2 years of Numerical Weather Prediction (NWP) data from Central Weather Bureau (CWB). Furthermore, forecasting results show that the proposed model outperforms other models such as Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) in terms of NRMSE and NMAPE.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134274904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774475
Zhi-Wei Liu, Shen-Li Chen, Jhong-Yi Lai, Yu-Jie Chung, Xing-Chen Mai
In this paper, a parasitic silicon controlled rectifier (SCR) area modulation at the drain-side of high-voltage n-type laterally-diffused metal-oxide semiconductor(nLDMOS) is proposed to evaluate the change of electrostatic discharge+ (ESD) capability with the increasing of SCR area. As the P doping area of the drain side is larger, the SCR characteristics are more obvious. Therefore, it can be found from the experimental data that when the P+ doping area is 100% (the nLD_SCR_P100), the It2 value of the device will increase from 1.38 A of the reference device to 9 A (an increase of 652.1%). Although the Vt1 and Vh values of the device also decreaseIt2, slightly as the SCR area becomes larger, but it has the best so the FOM increases from 1659.13 µA×V/µm2 for the reference device to 8098.56 µA×V/µm2(an increase of 488.1%), the ability to conduct ESD current is significantly improved.
{"title":"HV 65V nLDMOSs Engineering of ESD Enhancement by the Drain Side with Parasitic SCR Modulations","authors":"Zhi-Wei Liu, Shen-Li Chen, Jhong-Yi Lai, Yu-Jie Chung, Xing-Chen Mai","doi":"10.1109/ICASI55125.2022.9774475","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774475","url":null,"abstract":"In this paper, a parasitic silicon controlled rectifier (SCR) area modulation at the drain-side of high-voltage n-type laterally-diffused metal-oxide semiconductor(nLDMOS) is proposed to evaluate the change of electrostatic discharge+ (ESD) capability with the increasing of SCR area. As the P doping area of the drain side is larger, the SCR characteristics are more obvious. Therefore, it can be found from the experimental data that when the P+ doping area is 100% (the nLD_SCR_P100), the It2 value of the device will increase from 1.38 A of the reference device to 9 A (an increase of 652.1%). Although the Vt1 and Vh values of the device also decreaseIt2, slightly as the SCR area becomes larger, but it has the best so the FOM increases from 1659.13 µA×V/µm2 for the reference device to 8098.56 µA×V/µm2(an increase of 488.1%), the ability to conduct ESD current is significantly improved.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"183 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116064279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/icasi55125.2022.9774457
{"title":"ICASI 2022 TOC","authors":"","doi":"10.1109/icasi55125.2022.9774457","DOIUrl":"https://doi.org/10.1109/icasi55125.2022.9774457","url":null,"abstract":"","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126088223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774478
Yu-En Wu
This paper proposes a novel emergency lighting device (ELD), it can be used as an auxiliary lighting equipment to improve the utilization rate of equipment. Two kinds of topologies are used in ELD to achieve the functions of charging the battery and lighting. The battery charging topology uses a two-switch forward, with three-stage charging control to charge the battery when the AC is normal; the lighting circuit uses an improved boost-flyback converter. The output of the lighting circuit will be connected to the Relay. When the external power is stopped or the signal is turned on, the LED light is illuminated. The ELD first tests the functional stability through the burning machine, and then goes to the actual field for three months to test the overall stability of the circuit. The maximum efficiency of battery charging circuit reaches above 95% at a 75% load condition and the maximum efficiency of LED lighting circuit reaches above 98% at a 75% load condition.
{"title":"Development of a Novel Emergency Lighting System with Improved Boost-Flyback Topology","authors":"Yu-En Wu","doi":"10.1109/ICASI55125.2022.9774478","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774478","url":null,"abstract":"This paper proposes a novel emergency lighting device (ELD), it can be used as an auxiliary lighting equipment to improve the utilization rate of equipment. Two kinds of topologies are used in ELD to achieve the functions of charging the battery and lighting. The battery charging topology uses a two-switch forward, with three-stage charging control to charge the battery when the AC is normal; the lighting circuit uses an improved boost-flyback converter. The output of the lighting circuit will be connected to the Relay. When the external power is stopped or the signal is turned on, the LED light is illuminated. The ELD first tests the functional stability through the burning machine, and then goes to the actual field for three months to test the overall stability of the circuit. The maximum efficiency of battery charging circuit reaches above 95% at a 75% load condition and the maximum efficiency of LED lighting circuit reaches above 98% at a 75% load condition.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130113417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774460
Ming-Hong Hsieh, Chun-Chien Hsieh, Jia-Wun Wu, Po-Tsung Tseng, Kuang-Pang Lin
The fire and explosion hazards often caused by electrostatic discharge (ESD) especially in the glue coating operation of adhesive tape process, therefore the ESD protective measure is critical during tape manufacturing. Let the charged tape contact conductive material and connected to earth to prevent the charges accumulation is commonly employed in tape industry, and ESD fabric is one of the commonly used conductive materials. This study focus on the effectiveness of grounding by ESD fabric, and all fabric came from tape production company.The average resistivity of vertical line of ESD fabric is 5 Ω， the transverse line and band is 17GΩ same as the diagonal line and band of both side. In actual ESD protection settings, the contact points of the grounding system of the ESD fabric bond on both sides of the textile. Thus; the efficiency of earthing is poor due to the high resistivity of fabric. We covered the ground side (less than 40% of whole area) of the cloth with aluminum foil to ameliorate efficiency of earthing. After the improvement; the average resistivity of the transverse band on aluminum foil is less than 1Ω，the transverse band without aluminum foil covered from 17GΩ reduce to 381Ω， and diagonal band of both sides drop to less than 65Ω. According to previous consequence; the conductive performance of the grounding of the ESD fabric is obviously improved, and the research results are also spread to the tape industry.
{"title":"To Improve The Efficiency of Electrostatic Discharge Protection Measures for Glue Coating Operation of Adhesive Tape Process_ As an Example of Esd Fabric","authors":"Ming-Hong Hsieh, Chun-Chien Hsieh, Jia-Wun Wu, Po-Tsung Tseng, Kuang-Pang Lin","doi":"10.1109/ICASI55125.2022.9774460","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774460","url":null,"abstract":"The fire and explosion hazards often caused by electrostatic discharge (ESD) especially in the glue coating operation of adhesive tape process, therefore the ESD protective measure is critical during tape manufacturing. Let the charged tape contact conductive material and connected to earth to prevent the charges accumulation is commonly employed in tape industry, and ESD fabric is one of the commonly used conductive materials. This study focus on the effectiveness of grounding by ESD fabric, and all fabric came from tape production company.The average resistivity of vertical line of ESD fabric is 5 Ω， the transverse line and band is 17GΩ same as the diagonal line and band of both side. In actual ESD protection settings, the contact points of the grounding system of the ESD fabric bond on both sides of the textile. Thus; the efficiency of earthing is poor due to the high resistivity of fabric. We covered the ground side (less than 40% of whole area) of the cloth with aluminum foil to ameliorate efficiency of earthing. After the improvement; the average resistivity of the transverse band on aluminum foil is less than 1Ω，the transverse band without aluminum foil covered from 17GΩ reduce to 381Ω， and diagonal band of both sides drop to less than 65Ω. According to previous consequence; the conductive performance of the grounding of the ESD fabric is obviously improved, and the research results are also spread to the tape industry.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129750276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to solve the problem of hydrogen storage and transportation, the first thought is to store the hydrogen directly in the cylinder. However, this approach is not advisable, because even under the high pressure of 700 atm, the weight of hydrogen in the cylinder is less than the cylinder itself 1% of it, and it is very unsafe. Once it leaks, it will explode when it encounters sparks. In this study, we designed a lightweight hydrogen storage/release bottle. The hydrogen storage bottle is made of aluminum alloy. This hydrogen storage alloy in the hydrogen storage/release bottle is used to adsorb hydrogen, and the gas valve of the bottle is used to release hydrogen. In order to achieve the purpose of light weight, this portable hydrogen storage bottle was made of 7075 aluminum alloy. After the bottle body is hollowed out by machining, an integrally formed hydrogen storage bottle body is formed. The aluminum gasket is fixed to the flange structure, so the overall hydrogen storage bottle has no welded part. Further, a hard anodic films were formed on the bottle inner an d outer surfaces to increase the tensile stress of the bottle, prevent hydrogen embrittlement in the bottle, and anti-corrosion on the bottle. The design of this hydrogen storage tank includes the following features: (1) the lightweight portable storage/desorption of aluminum alloy bottle, (2) The outer layer of the bottle body is covered with an anodic film to strengthen bottle tension, (3) The outside of the bottle with a high quality anodic film, and (4) the inside of the bottle is anodized to prevent hydrogen embrittlement.
{"title":"Lightweight Portable and High Thermal Conductivity Storage/Release Hydrogen Bottle","authors":"Chao-Yen Chen, Xi-Wen Yang, Jun-Hong Lai, Chien-Chon Chen","doi":"10.1109/ICASI55125.2022.9774489","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774489","url":null,"abstract":"In order to solve the problem of hydrogen storage and transportation, the first thought is to store the hydrogen directly in the cylinder. However, this approach is not advisable, because even under the high pressure of 700 atm, the weight of hydrogen in the cylinder is less than the cylinder itself 1% of it, and it is very unsafe. Once it leaks, it will explode when it encounters sparks. In this study, we designed a lightweight hydrogen storage/release bottle. The hydrogen storage bottle is made of aluminum alloy. This hydrogen storage alloy in the hydrogen storage/release bottle is used to adsorb hydrogen, and the gas valve of the bottle is used to release hydrogen. In order to achieve the purpose of light weight, this portable hydrogen storage bottle was made of 7075 aluminum alloy. After the bottle body is hollowed out by machining, an integrally formed hydrogen storage bottle body is formed. The aluminum gasket is fixed to the flange structure, so the overall hydrogen storage bottle has no welded part. Further, a hard anodic films were formed on the bottle inner an d outer surfaces to increase the tensile stress of the bottle, prevent hydrogen embrittlement in the bottle, and anti-corrosion on the bottle. The design of this hydrogen storage tank includes the following features: (1) the lightweight portable storage/desorption of aluminum alloy bottle, (2) The outer layer of the bottle body is covered with an anodic film to strengthen bottle tension, (3) The outside of the bottle with a high quality anodic film, and (4) the inside of the bottle is anodized to prevent hydrogen embrittlement.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129849947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774441
Chun-Yu Chen, Wei-Jen Wang, Chun-Chuan Chen
Multiclass Classification of EEG signal is essential for brain computer interface (BCI) applications but extremely time consuming. We proposed a subject-weighted adaptive transfer learning method in conjunction with MLP and CNN classifiers for fast classification of multiclass EEG dataset.Analytic results show that CNN generally outperforms MLP in this multi-class classification. The use of transfer learning is efficient for building the predictive model without decreasing the accuracy and 2D CNN is more robust to between-subject variabilities.
{"title":"Multiclass Classification of EEG Motor Imagery Signals Based on Transfer Learning","authors":"Chun-Yu Chen, Wei-Jen Wang, Chun-Chuan Chen","doi":"10.1109/ICASI55125.2022.9774441","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774441","url":null,"abstract":"Multiclass Classification of EEG signal is essential for brain computer interface (BCI) applications but extremely time consuming. We proposed a subject-weighted adaptive transfer learning method in conjunction with MLP and CNN classifiers for fast classification of multiclass EEG dataset.Analytic results show that CNN generally outperforms MLP in this multi-class classification. The use of transfer learning is efficient for building the predictive model without decreasing the accuracy and 2D CNN is more robust to between-subject variabilities.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128961808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The extremely high transmission rate of the COVID-19 has made the supply of medical resources in countries around the world in short supply. The implementation of quarantine in order to avoid group infections has a serious impact on the economy, transportation, education and other aspects. Epidemic prevention will be a routine task that needs to be carried out for a long time and cannot be neglected. In view of the fact that wearing masks is currently an effective method of epidemic prevention, and the current face detection models are not effective for masked faces, and pedestrians who have not worn masks in the correct way. It may spread the epidemic. This research will establish a face data set with three kinds of annotations, and combine a variety of deep learning convolutional neural network architectures and methods to design a face detection model that can quickly train and detect wearing a mask, not wearing a mask, and wearing a mask incorrectly faces. In the hope of contributing to the epidemic prevention, we use an adaptive algorithm to adjust the image size to reduce unnecessary operations, and modify the CIOU_LOSS error function to speed up the operation. Experiments have confirmed that our algorithm saves 70% of the time compared to YOLO v5m with the same accuracy.
{"title":"A Novel Facial Mask Detection Using Fast-YOLO Algorithm","authors":"Chih-Chen Liu, Su-Chi Fuh, Chen-Jie Lin, Tzu-Hua Huang","doi":"10.1109/ICASI55125.2022.9774466","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774466","url":null,"abstract":"The extremely high transmission rate of the COVID-19 has made the supply of medical resources in countries around the world in short supply. The implementation of quarantine in order to avoid group infections has a serious impact on the economy, transportation, education and other aspects. Epidemic prevention will be a routine task that needs to be carried out for a long time and cannot be neglected. In view of the fact that wearing masks is currently an effective method of epidemic prevention, and the current face detection models are not effective for masked faces, and pedestrians who have not worn masks in the correct way. It may spread the epidemic. This research will establish a face data set with three kinds of annotations, and combine a variety of deep learning convolutional neural network architectures and methods to design a face detection model that can quickly train and detect wearing a mask, not wearing a mask, and wearing a mask incorrectly faces. In the hope of contributing to the epidemic prevention, we use an adaptive algorithm to adjust the image size to reduce unnecessary operations, and modify the CIOU_LOSS error function to speed up the operation. Experiments have confirmed that our algorithm saves 70% of the time compared to YOLO v5m with the same accuracy.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129218296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774449
Chih-Feng Yen, H. Hsu, Yu-Ya Huang, Shen-Hao Tsao
In this study, zinc oxide (ZnO) doped yttrium (Y2O3) thin film was successfully prepared by liquid phase deposition (LPD). The samples annealed in different gas environments is used to observe the relationship between annealing environment and electrical characteristics. We found that samples annealed in air have the best electrical properties. Mixing different density of ZnO to Y2O3 shows that ZnO doped with 25 % has the best electrical characteristics. The oxide capacitance (Cox), k value, interface trap level density (Dit) and the leakage current density are 248.22 pF, 31.9, 1.08 × 1012 eV-1cm-2 and 5.62 × 10-6 A/cm2 at + 5 V, respectively.
{"title":"Study on Y2O3 based gate oxide Si MOS with various annealing ambient","authors":"Chih-Feng Yen, H. Hsu, Yu-Ya Huang, Shen-Hao Tsao","doi":"10.1109/ICASI55125.2022.9774449","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774449","url":null,"abstract":"In this study, zinc oxide (ZnO) doped yttrium (Y<inf>2</inf>O<inf>3</inf>) thin film was successfully prepared by liquid phase deposition (LPD). The samples annealed in different gas environments is used to observe the relationship between annealing environment and electrical characteristics. We found that samples annealed in air have the best electrical properties. Mixing different density of ZnO to Y<inf>2</inf>O<inf>3</inf> shows that ZnO doped with 25 % has the best electrical characteristics. The oxide capacitance (C<inf>ox</inf>), k value, interface trap level density (D<inf>it</inf>) and the leakage current density are 248.22 pF, 31.9, 1.08 × 10<sup>12</sup> eV<sup>-1</sup>cm<sup>-2</sup> and 5.62 × 10<sup>-6</sup> A/cm<sup>2</sup> at + 5 V, respectively.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131159506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-22DOI: 10.1109/ICASI55125.2022.9774468
Chenguang Li, C. Su
In this paper, we propose a different method of deepening the network model of YOLOv5s6 and design three types of Multi-Connection (MC) blocks that are suitable for specific datasets. The main purpose of Multi-Connection block is to reuse features and retain input features for passing down. Eight public datasets and one customized dataset are experimented for verification. We improve the residual block in YOLOv5. The results show that the average precision (AP) can be increased. Compared with the YOLOv5s6, YOLOv5s6 with MC type I increases the AP by 1.6% in the Global Wheat Head Dataset 2020, YOLOv5s6 with MC type II increases the AP by 2.9% in the PlanDoc dataset, and YOLOv5s6 with MC type III increases the AP by 2.9% in the PASCAL Visual Object Classes (VOC) dataset. Using the multi-connection of double residual block performs better than the original residual block of YOLOv5s6.
{"title":"Multi-Connection of Double Residual Block for YOLOv5 Object Detection","authors":"Chenguang Li, C. Su","doi":"10.1109/ICASI55125.2022.9774468","DOIUrl":"https://doi.org/10.1109/ICASI55125.2022.9774468","url":null,"abstract":"In this paper, we propose a different method of deepening the network model of YOLOv5s6 and design three types of Multi-Connection (MC) blocks that are suitable for specific datasets. The main purpose of Multi-Connection block is to reuse features and retain input features for passing down. Eight public datasets and one customized dataset are experimented for verification. We improve the residual block in YOLOv5. The results show that the average precision (AP) can be increased. Compared with the YOLOv5s6, YOLOv5s6 with MC type I increases the AP by 1.6% in the Global Wheat Head Dataset 2020, YOLOv5s6 with MC type II increases the AP by 2.9% in the PlanDoc dataset, and YOLOv5s6 with MC type III increases the AP by 2.9% in the PASCAL Visual Object Classes (VOC) dataset. Using the multi-connection of double residual block performs better than the original residual block of YOLOv5s6.","PeriodicalId":190229,"journal":{"name":"2022 8th International Conference on Applied System Innovation (ICASI)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124469437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：