2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)最新文献
Pub Date : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704918
H. Nieto-Chaupis
In this paper a mathematical model that focuses at the very beginning of pandemic at Europe is presented. In essence it is assumed that once the virus arrived to Italy then the geographical propagation was done through probabilistic rules among then to Spain. Because of this the model of propagation of Feynman in conjunction to Wiener schemes have been used to model the displacement of virus from Wuhan to Milan as well from Milan to Spain, as seen at the end of 2019 triggering the beginning of European pandemic at January of 2020. As seen at official data Italy and Spain have presented same statistics at the first months of local pandemic. From the usage of the proposed formalism, it is found that the country data are following Gaussian-like distributions due to the space-time propagation of virus.
{"title":"Spreading of Corona Virus in Central Europe: The Cases of Italy and Spain and the Very Beginning of Pandemic","authors":"H. Nieto-Chaupis","doi":"10.1109/SNPD51163.2021.9704918","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704918","url":null,"abstract":"In this paper a mathematical model that focuses at the very beginning of pandemic at Europe is presented. In essence it is assumed that once the virus arrived to Italy then the geographical propagation was done through probabilistic rules among then to Spain. Because of this the model of propagation of Feynman in conjunction to Wiener schemes have been used to model the displacement of virus from Wuhan to Milan as well from Milan to Spain, as seen at the end of 2019 triggering the beginning of European pandemic at January of 2020. As seen at official data Italy and Spain have presented same statistics at the first months of local pandemic. From the usage of the proposed formalism, it is found that the country data are following Gaussian-like distributions due to the space-time propagation of virus.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123293469","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9705014
Menq-Jiun Wu, Ye Chen, Yi-Sheng Liao, Jun-An Chen, Hao-Han Lin
The theory of face recognition is mainly from the idea of feature vectors. The face image is converted into a series of numbers to form a feature vector. For a feature vector of a face image, its content includes various features, such as: face height, face width, average face color, lips width, and nose height. The face-recognition operation is to compare the input of the feature vector of a face image with a large number of feature vectors in a dataset to identify the personal identity.The face recognition system in this paper is mainly implemented in the Python environment. Face generation is achieved by selfie of face. The image is cut to retain the part of the face, and stored in the database. Comparing the face image input with those saved in the dataset, if the similarity value passes the threshold value of true. The program will show the face image identification. Otherwise, the system will display a false message. The face recognition system is completed and the experimental results show the correct face-recognition. Finally the laptop Webcam is used to take the face image, and the result of comparison is also correct.
{"title":"Face-recognition System Design and Manufacture","authors":"Menq-Jiun Wu, Ye Chen, Yi-Sheng Liao, Jun-An Chen, Hao-Han Lin","doi":"10.1109/SNPD51163.2021.9705014","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9705014","url":null,"abstract":"The theory of face recognition is mainly from the idea of feature vectors. The face image is converted into a series of numbers to form a feature vector. For a feature vector of a face image, its content includes various features, such as: face height, face width, average face color, lips width, and nose height. The face-recognition operation is to compare the input of the feature vector of a face image with a large number of feature vectors in a dataset to identify the personal identity.The face recognition system in this paper is mainly implemented in the Python environment. Face generation is achieved by selfie of face. The image is cut to retain the part of the face, and stored in the database. Comparing the face image input with those saved in the dataset, if the similarity value passes the threshold value of true. The program will show the face image identification. Otherwise, the system will display a false message. The face recognition system is completed and the experimental results show the correct face-recognition. Finally the laptop Webcam is used to take the face image, and the result of comparison is also correct.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125328181","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704932
Toshihiro Nakamura, Akito Monden, Mariko Sasakura, H. Uwano
Explaining a program to others helps get others to find bugs and for the explainer him/herself to find bugs. However, to the best of our knowledge, there is no quantitative evidence that explaining a program to others helps the explainer find bugs. This study aims to show quantitatively, using an experimental evaluation, that the explainer himself can find new bugs by explaining the program to others. In the experiment, subjects first review a program that contains many bugs and try to find as many bugs as possible. Next, they are required to explain the program aloud to others. We see if they notice any new bugs themselves during the explanation. As a result of the experiment, five out of the six subjects could find new bugs when explaining the program to others. According to the questionnaire to the subjects, the subjects who find many bugs feel that they can understand the program better by explaining it to others.
{"title":"Effectiveness of Explaining a Program to Others in Finding Its Bugs","authors":"Toshihiro Nakamura, Akito Monden, Mariko Sasakura, H. Uwano","doi":"10.1109/SNPD51163.2021.9704932","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704932","url":null,"abstract":"Explaining a program to others helps get others to find bugs and for the explainer him/herself to find bugs. However, to the best of our knowledge, there is no quantitative evidence that explaining a program to others helps the explainer find bugs. This study aims to show quantitatively, using an experimental evaluation, that the explainer himself can find new bugs by explaining the program to others. In the experiment, subjects first review a program that contains many bugs and try to find as many bugs as possible. Next, they are required to explain the program aloud to others. We see if they notice any new bugs themselves during the explanation. As a result of the experiment, five out of the six subjects could find new bugs when explaining the program to others. According to the questionnaire to the subjects, the subjects who find many bugs feel that they can understand the program better by explaining it to others.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116416917","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704975
W. Chang, Rong-Terng Juang, Min-Hsiang Huang, M. Sung
Vehicles are one kind of human’s transport tools that can carry us to anywhere. Safety is an important issue for vehicle applications. Tire is an important element of the vehicle, because the vehicle only has four points connect on the road through tires. Many accidents are caused the tire problems. There are more and more researches in intelligent tire base on safety issues. Tire pressure monitor system (TPMS) can be called an application of intelligent tires that are currently a very popular tire monitoring device, it provides pressure and temperature information for drivers. The information can let drivers understand their tire’s status through simple information. Not only pressure and temperature can affect safety, but also tire’s use mileage. The tire wear has some relation to safety. The mileage has been defined by tire spec. The different tires have different specs and the tire wear status cannot be defined clearly. The wear status is hard to get that is because the vehicle condition, use scenario and situation are complicated. Drivers can’t understand how many miles did the tire use and what status the tire. This paper proposed a method to estimate and monitor tire mileage and wear base on simple sensors. G-sensor and hall sensors are common sensors in applications. The method will use these simple sensors to implement the method. Drivers can accord information to evaluate the tire status and decide the timing to replace the tire. It can help drivers to increase safety during the drive and provide a solution to support intelligent tire function.
{"title":"Implementation of Tire Status Estimation Using Hall Sensor and G-Sensor","authors":"W. Chang, Rong-Terng Juang, Min-Hsiang Huang, M. Sung","doi":"10.1109/SNPD51163.2021.9704975","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704975","url":null,"abstract":"Vehicles are one kind of human’s transport tools that can carry us to anywhere. Safety is an important issue for vehicle applications. Tire is an important element of the vehicle, because the vehicle only has four points connect on the road through tires. Many accidents are caused the tire problems. There are more and more researches in intelligent tire base on safety issues. Tire pressure monitor system (TPMS) can be called an application of intelligent tires that are currently a very popular tire monitoring device, it provides pressure and temperature information for drivers. The information can let drivers understand their tire’s status through simple information. Not only pressure and temperature can affect safety, but also tire’s use mileage. The tire wear has some relation to safety. The mileage has been defined by tire spec. The different tires have different specs and the tire wear status cannot be defined clearly. The wear status is hard to get that is because the vehicle condition, use scenario and situation are complicated. Drivers can’t understand how many miles did the tire use and what status the tire. This paper proposed a method to estimate and monitor tire mileage and wear base on simple sensors. G-sensor and hall sensors are common sensors in applications. The method will use these simple sensors to implement the method. Drivers can accord information to evaluate the tire status and decide the timing to replace the tire. It can help drivers to increase safety during the drive and provide a solution to support intelligent tire function.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128274342","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704947
Kunpeng Sun, Ji-Hwei Horng, C. Chang
Vector quantization (VQ) is a popular digital image compression technique. Its resulting index table can be further compressed using the side match vector quantization (SMVQ). In this research, we propose a reversible data hiding scheme based on the de-clustering rules to embed secret data during SMVQ compression. Referring to differently assigned codebooks, the de-clustering rules are equally applicable to both compressible and uncompressible VQ indices. The proposed scheme can produce a camouflaged VQ index table with a high payload. Besides, our scheme is free from the indicator bit, which is required in the conventional SMVQ. Experimental results are compared with state-of-the-art methods.
{"title":"Reversible Data Hiding for SMVQ Compressed Images Based on De-Clustering Rules","authors":"Kunpeng Sun, Ji-Hwei Horng, C. Chang","doi":"10.1109/SNPD51163.2021.9704947","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704947","url":null,"abstract":"Vector quantization (VQ) is a popular digital image compression technique. Its resulting index table can be further compressed using the side match vector quantization (SMVQ). In this research, we propose a reversible data hiding scheme based on the de-clustering rules to embed secret data during SMVQ compression. Referring to differently assigned codebooks, the de-clustering rules are equally applicable to both compressible and uncompressible VQ indices. The proposed scheme can produce a camouflaged VQ index table with a high payload. Besides, our scheme is free from the indicator bit, which is required in the conventional SMVQ. Experimental results are compared with state-of-the-art methods.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130204352","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704891
Hiroto Mizutani, Masateru Tsunoda, K. Nakasai
Background: Machine learning models are sometimes embedded in software to implement the required functions. As a result, non-experts in machine learning are becoming familiar with the models. However, the interpretability of the built models is often low in machine learning, such as deep learning, and the recognition process of such models is very different from that of humans. Therefore, it is not easy for novice users, such as end-users and beginners, to anticipate the behavior of models that they will use or build. Aim: We assist novice users to realize an aspect of the behavior of machine learning models relating to robustness intuitively. Method: We formalized and evaluated quiz-based analysis, which is often applied by practitioners to test the robustness of machine learning models arbitrarily. To generate test cases of the models, the analysis converts images towards the boundary of classification for both machine learning and humans. It can be regarded as a type of boundary value analysis of software development. Results: In the experiment, we evaluated whether the analysis quantitatively clarified the aspects of the models. The analysis clarified the robustness of the model for image conversion and misclassification quantitatively. Conclusion: The analysis is expected to enlighten novice users on the behavior of machine learning models. This may promote behavioral changes in the evaluation of models for novice users.
{"title":"How to Enlighten Novice Users on Behavior of Machine Learning Models?","authors":"Hiroto Mizutani, Masateru Tsunoda, K. Nakasai","doi":"10.1109/SNPD51163.2021.9704891","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704891","url":null,"abstract":"Background: Machine learning models are sometimes embedded in software to implement the required functions. As a result, non-experts in machine learning are becoming familiar with the models. However, the interpretability of the built models is often low in machine learning, such as deep learning, and the recognition process of such models is very different from that of humans. Therefore, it is not easy for novice users, such as end-users and beginners, to anticipate the behavior of models that they will use or build. Aim: We assist novice users to realize an aspect of the behavior of machine learning models relating to robustness intuitively. Method: We formalized and evaluated quiz-based analysis, which is often applied by practitioners to test the robustness of machine learning models arbitrarily. To generate test cases of the models, the analysis converts images towards the boundary of classification for both machine learning and humans. It can be regarded as a type of boundary value analysis of software development. Results: In the experiment, we evaluated whether the analysis quantitatively clarified the aspects of the models. The analysis clarified the robustness of the model for image conversion and misclassification quantitatively. Conclusion: The analysis is expected to enlighten novice users on the behavior of machine learning models. This may promote behavioral changes in the evaluation of models for novice users.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114796269","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704991
Chinchen Chang, T. Nguyen, Chia-Chen Lin
In this paper, we present our design of a new, blind, reversible, robust, watermarking scheme for relational databases storing textual and numerical data, which is not only guaranteed to meet the above seven requirements and to provide proof of ownership for the legal owner of a relational database but also does not require the existence of the primary key attribute in the relational database. Extensive analysis and our experimental results showed that the proposed scheme achieves strong robustness under a variety of possible attacks, e.g., deletion attacks, modification attacks, addition attacks, and sorting attacks. In addition, in comparison with six existing schemes, the proposed scheme was more secure and robust, as evidenced by our experimental results and robustness analysis.
{"title":"A Reversible Database Watermark Scheme for Textual and Numerical Datasets","authors":"Chinchen Chang, T. Nguyen, Chia-Chen Lin","doi":"10.1109/SNPD51163.2021.9704991","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704991","url":null,"abstract":"In this paper, we present our design of a new, blind, reversible, robust, watermarking scheme for relational databases storing textual and numerical data, which is not only guaranteed to meet the above seven requirements and to provide proof of ownership for the legal owner of a relational database but also does not require the existence of the primary key attribute in the relational database. Extensive analysis and our experimental results showed that the proposed scheme achieves strong robustness under a variety of possible attacks, e.g., deletion attacks, modification attacks, addition attacks, and sorting attacks. In addition, in comparison with six existing schemes, the proposed scheme was more secure and robust, as evidenced by our experimental results and robustness analysis.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124797161","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9704981
H. Nieto-Chaupis
At the end of first quarter of 2020 it was seen in most countries statistics the beginning of an imminent second wave of pandemic. On January of 2021 it was seen in the data a rapid growth of new infections. In this paper, a geometry-based scheme is presented. In concrete the rectangle and trapezoid shapes are analyzed. From this, a relation between both geometries is extracted in terms of polynomial functions. The resulting characterization of a pandemic in terms of geometric variables is presented. Thus the present model is confronted with official data of USA and India. From the results of this paper, it is strongly believed that entropy might be behind of a global pandemic dynamics.
{"title":"Geometrical Schemes as Probabilistic and Entropic Tools to Estimate Duration and Peaks of Pandemic Waves","authors":"H. Nieto-Chaupis","doi":"10.1109/SNPD51163.2021.9704981","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704981","url":null,"abstract":"At the end of first quarter of 2020 it was seen in most countries statistics the beginning of an imminent second wave of pandemic. On January of 2021 it was seen in the data a rapid growth of new infections. In this paper, a geometry-based scheme is presented. In concrete the rectangle and trapezoid shapes are analyzed. From this, a relation between both geometries is extracted in terms of polynomial functions. The resulting characterization of a pandemic in terms of geometric variables is presented. Thus the present model is confronted with official data of USA and India. From the results of this paper, it is strongly believed that entropy might be behind of a global pandemic dynamics.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116085417","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}
Traumatic Brain Injury (TBI) is most commonly accident injury in modern society. The deterioration of cognition and memory is a common phenomenon caused by TBI. Most of the basic experiments used rats for pathological research. Moreover, an eight-arm maze was often used to test the spatial learning behavior of brain diseases, such as Alzheimer’s disease and TBI. However, most of maze experimental data were collected in manual records. This process would take a lot of manpower and time. Therefore, the research built an automatic tracking trajectory system of the eight-arm maze to collect experiment data. Furthermore, the path trajectory of the rat can be recorded in time. Finally, these path trajectory data were used to analysis behavior feature of TBI animals. The results showed that TBI rats have a 40%~80% chance of having a trajectory to the right.
{"title":"Automated Eight-Arm Maze Trajectory Tracking System for Feature Extraction of TBI Animals","authors":"Shu-Cing Wu, Chi-Yuan Lin, Liang-Jyun Hong, Chi-Chun Chen","doi":"10.1109/SNPD51163.2021.9704966","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9704966","url":null,"abstract":"Traumatic Brain Injury (TBI) is most commonly accident injury in modern society. The deterioration of cognition and memory is a common phenomenon caused by TBI. Most of the basic experiments used rats for pathological research. Moreover, an eight-arm maze was often used to test the spatial learning behavior of brain diseases, such as Alzheimer’s disease and TBI. However, most of maze experimental data were collected in manual records. This process would take a lot of manpower and time. Therefore, the research built an automatic tracking trajectory system of the eight-arm maze to collect experiment data. Furthermore, the path trajectory of the rat can be recorded in time. Finally, these path trajectory data were used to analysis behavior feature of TBI animals. The results showed that TBI rats have a 40%~80% chance of having a trajectory to the right.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121702801","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 : 2021-11-24DOI: 10.1109/SNPD51163.2021.9705011
Zheng-An Zhu, Chien-Hao Chen, Chen-Kuo Chiang
Outdoor scene parsing is a very popular topic which algorithms seek to labels or identify objects in images. Sky segmentation is one of the popular outdoor scene parsing task. Sky segmentation models are usually trained on ideal datasets and produce high quality results. However, the performance of sky segmentation model decreases because of varying weather conditions, different time and scene changes due to seasonal weather or other issues in reality. This paper focuses on applying data augmentation methods to generate diversified images. A conditional data augmentation method based on BicycleGAN is proposed in this paper. The model considers mask loss and content loss for improving the quality and details of the generated images. The experimental results demonstrate that the quality of the generated image is better than the existing methods.
{"title":"Conditional Data Augmentation For Sky Segmentation","authors":"Zheng-An Zhu, Chien-Hao Chen, Chen-Kuo Chiang","doi":"10.1109/SNPD51163.2021.9705011","DOIUrl":"https://doi.org/10.1109/SNPD51163.2021.9705011","url":null,"abstract":"Outdoor scene parsing is a very popular topic which algorithms seek to labels or identify objects in images. Sky segmentation is one of the popular outdoor scene parsing task. Sky segmentation models are usually trained on ideal datasets and produce high quality results. However, the performance of sky segmentation model decreases because of varying weather conditions, different time and scene changes due to seasonal weather or other issues in reality. This paper focuses on applying data augmentation methods to generate diversified images. A conditional data augmentation method based on BicycleGAN is proposed in this paper. The model considers mask loss and content loss for improving the quality and details of the generated images. The experimental results demonstrate that the quality of the generated image is better than the existing methods.","PeriodicalId":235370,"journal":{"name":"2021 IEEE/ACIS 22nd International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-11-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121833520","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：