Pub Date : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179494
Zongcheng Yue, Chiu-Wing Sham, C. Y. Lo, W. Cheung, C. Yiu
Semantic segmentation in computer vision is a challenging area of research, aiming to accurately segment and categorize objects and regions within an image. One widely used dataset for this task is Cityscapes, which contains a variety of city-related object classes such as cars, pedestrians, bicycles, and buildings. However, the Cityscapes dataset does not include any aquatic view classes, which limits its potential for applications in coastal and marine environments. This paper presents a novel approach to extend the Cityscapes dataset with aquatic classes to address this limitation. Our proposed method involves the implementation of two state-of-the-art neural network models, one based on the Cityscapes dataset and the other on a common aquatic dataset. We then selectively extract the aquatic segmen-tation results from the corresponding model according to the aquatic label. We further generate a mask image for the sea class and merge it precisely with the resulting image from the Cityscapes-based model. Our method is evaluated by comparing the performance of the original Cityscapes-based model with the extended Cityscapes-based model on a set of test images that contain aquatic views. The results show that our approach can maintain the original model’s high segmentation accuracy for all views except for aquatic areas while preserving the relevant parts of the marine model in terms of accuracy and area coverage. Additionally, our approach does not require retraining, thus saving computational resources and time.
{"title":"Sea View Extension for Semantic Segmentation in Cityscapes","authors":"Zongcheng Yue, Chiu-Wing Sham, C. Y. Lo, W. Cheung, C. Yiu","doi":"10.1109/ICASI57738.2023.10179494","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179494","url":null,"abstract":"Semantic segmentation in computer vision is a challenging area of research, aiming to accurately segment and categorize objects and regions within an image. One widely used dataset for this task is Cityscapes, which contains a variety of city-related object classes such as cars, pedestrians, bicycles, and buildings. However, the Cityscapes dataset does not include any aquatic view classes, which limits its potential for applications in coastal and marine environments. This paper presents a novel approach to extend the Cityscapes dataset with aquatic classes to address this limitation. Our proposed method involves the implementation of two state-of-the-art neural network models, one based on the Cityscapes dataset and the other on a common aquatic dataset. We then selectively extract the aquatic segmen-tation results from the corresponding model according to the aquatic label. We further generate a mask image for the sea class and merge it precisely with the resulting image from the Cityscapes-based model. Our method is evaluated by comparing the performance of the original Cityscapes-based model with the extended Cityscapes-based model on a set of test images that contain aquatic views. The results show that our approach can maintain the original model’s high segmentation accuracy for all views except for aquatic areas while preserving the relevant parts of the marine model in terms of accuracy and area coverage. Additionally, our approach does not require retraining, thus saving computational resources and time.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122257513","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179541
Jia-Jhe Song, Wei-Jen Chen, Yung-Fang Chen, S. Tseng
Previously, we proposed schemes in [1] and [2] for the classical subcarrier, bit, and power allocation problem [3] to minimize the total transmit power for multiuser orthogonal frequency division multiplexing systems in downlink transmission. In this paper, we propose a deep neural network (DNN) structure to speed up solving this complex problem. We propose a deep learning frame structure in which each group of allocation is termed as a batch; after some numbers of iterations and epochs, the loss will tend to converge to a constant value. The simulation results reveal that the proposed DNN-based schemes offer competitive performance and reduce computing time tremendously compared with those of the existing approaches.
{"title":"Subcarrier Allocation for Multiuser OFDM Systems by Using Deep Neural Networks","authors":"Jia-Jhe Song, Wei-Jen Chen, Yung-Fang Chen, S. Tseng","doi":"10.1109/ICASI57738.2023.10179541","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179541","url":null,"abstract":"Previously, we proposed schemes in [1] and [2] for the classical subcarrier, bit, and power allocation problem [3] to minimize the total transmit power for multiuser orthogonal frequency division multiplexing systems in downlink transmission. In this paper, we propose a deep neural network (DNN) structure to speed up solving this complex problem. We propose a deep learning frame structure in which each group of allocation is termed as a batch; after some numbers of iterations and epochs, the loss will tend to converge to a constant value. The simulation results reveal that the proposed DNN-based schemes offer competitive performance and reduce computing time tremendously compared with those of the existing approaches.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132247387","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179595
Trisha Jane Tejones, Angelito A. Silverio, Rodney M. Manalo
This study presents a double-tail latched comparator with a sub-mV input offset voltage. The Input Offset Storage (IOS) and transistor sizing techniques were used to address the two leading contributors of offset, which are the preamplifier’s input pair, and the regenerative latch’s reset switches. The proposed circuit was implemented using 180 nm CMOS technology, and operated under a 1. 8V supply and 10 kHz clock frequency. Results show that the combination of techniques leads to a total input offset voltage of942 uV, which is equivalent to a 75% reduction with respect to the original configuration. Usage of the IOS technique results in a static power dissipation of 19.82 uW. This translates to an average power of 83.8 nW when measured over 100 cycles, which is the interval before the IOS phase is repeated. Overall, the design exhibits superior robustness against the effects of mismatches, making it suitable for precision applications, such as biomedical data converters.
{"title":"A Comparator with sub-mV offset in Deep Submicron Technology for Biomedical Applications","authors":"Trisha Jane Tejones, Angelito A. Silverio, Rodney M. Manalo","doi":"10.1109/ICASI57738.2023.10179595","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179595","url":null,"abstract":"This study presents a double-tail latched comparator with a sub-mV input offset voltage. The Input Offset Storage (IOS) and transistor sizing techniques were used to address the two leading contributors of offset, which are the preamplifier’s input pair, and the regenerative latch’s reset switches. The proposed circuit was implemented using 180 nm CMOS technology, and operated under a 1. 8V supply and 10 kHz clock frequency. Results show that the combination of techniques leads to a total input offset voltage of942 uV, which is equivalent to a 75% reduction with respect to the original configuration. Usage of the IOS technique results in a static power dissipation of 19.82 uW. This translates to an average power of 83.8 nW when measured over 100 cycles, which is the interval before the IOS phase is repeated. Overall, the design exhibits superior robustness against the effects of mismatches, making it suitable for precision applications, such as biomedical data converters.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"T159 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125660886","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179599
Ruei-Hau Hsu, Bo Chen, Cheng-Jie Dai
As communication and high-speed internet make it easy to spread fake news on social media, scholars propose methods to detect it. However, existing approaches have limitations, such as reduced effectiveness without user information and high computational costs. Our proposed method, based on temporal and communication networks, is mainly used in the context of lack of user-related data and large textual datasets such as social media, forums, and online news. In sparse data settings, our proposed method can capture the propagation features of fake news for fake news detection, which is a feature extraction method based on building a propagation network for fake news detection. By studying the propagation pattern of fake news on social media, we obtain features belonging to the propagation network and test the source tweets using various machine learning classifiers. We also conduct experiments on realistic datasets to validate the method’s feasibility in social network scenarios.
{"title":"Intelligent Detection of Disinformation Based on Chronological and Spatial Topologies","authors":"Ruei-Hau Hsu, Bo Chen, Cheng-Jie Dai","doi":"10.1109/ICASI57738.2023.10179599","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179599","url":null,"abstract":"As communication and high-speed internet make it easy to spread fake news on social media, scholars propose methods to detect it. However, existing approaches have limitations, such as reduced effectiveness without user information and high computational costs. Our proposed method, based on temporal and communication networks, is mainly used in the context of lack of user-related data and large textual datasets such as social media, forums, and online news. In sparse data settings, our proposed method can capture the propagation features of fake news for fake news detection, which is a feature extraction method based on building a propagation network for fake news detection. By studying the propagation pattern of fake news on social media, we obtain features belonging to the propagation network and test the source tweets using various machine learning classifiers. We also conduct experiments on realistic datasets to validate the method’s feasibility in social network scenarios.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114039987","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179568
J. Liou, Zhi-Yu Lin, Zong-Xuan Hsieh
Most hemodialysis patients have blood oxygen levels of 95-100%. However, some people have blood oxygen levels below 95% and still lead normal lives. Slightly lower values while sleeping are normal and some users may measure values below 95%. Because each person’s condition is different, measure it every day, and record it before and after hemodialysis, so that the wearable wireless sensor can obtain the basic value change curve, so that there is a way to detect the problem.
{"title":"Blood Oxygen (SpO2) and Pulse Rate(PR) Wearable Sensor for Hemodialysis Patients","authors":"J. Liou, Zhi-Yu Lin, Zong-Xuan Hsieh","doi":"10.1109/ICASI57738.2023.10179568","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179568","url":null,"abstract":"Most hemodialysis patients have blood oxygen levels of 95-100%. However, some people have blood oxygen levels below 95% and still lead normal lives. Slightly lower values while sleeping are normal and some users may measure values below 95%. Because each person’s condition is different, measure it every day, and record it before and after hemodialysis, so that the wearable wireless sensor can obtain the basic value change curve, so that there is a way to detect the problem.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"6 5","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120817206","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179512
H. Hsu, Chen-Yu Pan
Indoor air quality reduces pollutants through different ventilation methods. Using different ventilation strategies is the focus of most scholars with limited resources. Therefore, we use outdoor environmental factors to data-driven control mechanical ventilation facilities.This proposed framework also optimizes the deep learning model (LSTM) through clustering analysis, and through cross-validation, the accuracy of the model is 97.45%. At the same time, this model can reduce energy consumption by 53%.
{"title":"Data-Driven Control of Mechanical Ventilation Using Open Data Environmental Factors","authors":"H. Hsu, Chen-Yu Pan","doi":"10.1109/ICASI57738.2023.10179512","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179512","url":null,"abstract":"Indoor air quality reduces pollutants through different ventilation methods. Using different ventilation strategies is the focus of most scholars with limited resources. Therefore, we use outdoor environmental factors to data-driven control mechanical ventilation facilities.This proposed framework also optimizes the deep learning model (LSTM) through clustering analysis, and through cross-validation, the accuracy of the model is 97.45%. At the same time, this model can reduce energy consumption by 53%.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"606 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116392156","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179597
I-Hui Li, L. Chang, You-Cheng Cao, Jian-Jun Sun, Xuan-You Liu, Guan-Lin Lu, Kun-Yi Cai
Since the impact of epidemic, people reduce the frequency of eating out and then cook by themselves. However, most people don’t know what to cook? And turn to the recipe for help, but they must check the existed ingredients in their refrigerator and compare one by one to find what to cook. Therefore, this research designed a smart refrigerator management system with recipe recommendation, ingredient management and smart reminder, to reduce cooking troubles, manage ingredients effectively and decrease food waste.
{"title":"A Smart Refrigerator Management System","authors":"I-Hui Li, L. Chang, You-Cheng Cao, Jian-Jun Sun, Xuan-You Liu, Guan-Lin Lu, Kun-Yi Cai","doi":"10.1109/ICASI57738.2023.10179597","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179597","url":null,"abstract":"Since the impact of epidemic, people reduce the frequency of eating out and then cook by themselves. However, most people don’t know what to cook? And turn to the recipe for help, but they must check the existed ingredients in their refrigerator and compare one by one to find what to cook. Therefore, this research designed a smart refrigerator management system with recipe recommendation, ingredient management and smart reminder, to reduce cooking troubles, manage ingredients effectively and decrease food waste.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"403 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116604209","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179524
Meng-Yi Lin, Jau‐Ji Jou, Chien-Liang Chiu, Chih-Yuan Lien, Bing-Hong Liu
Herein, we propose a 50 Gb/s non-return-to-zero optical modulator driver circuit in 0.18 μm SiGe BiCMOS technology. This modulator driver circuit including a pre-driver and a main driver was designed as a full-differential circuit. For the driver circuit, the bandwidth is 28.4 GHz, the voltage gain is 12.1 dB, the maximum differential output swing is 2 Vppd, and the chip area is 1.064➨1.067 mm2. The driver circuit can be applied in a high-speed micro-ring modulator with low driving voltage.
{"title":"A 50 Gb/s NRZ Optical Modulator Driver Circuit in 0.18 μm SiGe BiCMOS Technology","authors":"Meng-Yi Lin, Jau‐Ji Jou, Chien-Liang Chiu, Chih-Yuan Lien, Bing-Hong Liu","doi":"10.1109/ICASI57738.2023.10179524","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179524","url":null,"abstract":"Herein, we propose a 50 Gb/s non-return-to-zero optical modulator driver circuit in 0.18 μm SiGe BiCMOS technology. This modulator driver circuit including a pre-driver and a main driver was designed as a full-differential circuit. For the driver circuit, the bandwidth is 28.4 GHz, the voltage gain is 12.1 dB, the maximum differential output swing is 2 Vppd, and the chip area is 1.064➨1.067 mm2. The driver circuit can be applied in a high-speed micro-ring modulator with low driving voltage.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"151 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123389514","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179548
Hanseok Jeong, Jueun Jeong, Jonghoon Chun, Han-joon Kim
High-quality data can increase the reliability of machine learning-based prediction models. In our work, we propose a novel method for data correction to improve the quality of multivariate time-series data. For this, we use a LSTM-based VAE-GAN for anomaly detection and an Attention-based LSTM model for data correction. Through experiments using Secure Water Treatment (SWaT) data, we show that the proposed correction method is superior to previous correction methods.
{"title":"Multivariate Time-series Data Correction by combining Attention-based LSTM and GAN Model","authors":"Hanseok Jeong, Jueun Jeong, Jonghoon Chun, Han-joon Kim","doi":"10.1109/ICASI57738.2023.10179548","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179548","url":null,"abstract":"High-quality data can increase the reliability of machine learning-based prediction models. In our work, we propose a novel method for data correction to improve the quality of multivariate time-series data. For this, we use a LSTM-based VAE-GAN for anomaly detection and an Attention-based LSTM model for data correction. Through experiments using Secure Water Treatment (SWaT) data, we show that the proposed correction method is superior to previous correction methods.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121885525","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 : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179532
Chih-Hua Tai, Ya-Wen Teng
As social network platforms are becoming more and more popular, many activities spread event information through these platforms so that the number of participants during the event can reach the maximum. Note that as time goes by, the event information may be forgotten and repeatedly seen by the users. So the participation intention of a user for the event will change between active and inactive. Therefore, we formulate an influence maximization problem concerning the previous publicity and volatility of user behaviors given a specific period, and propose a fluctuation-aware independent-cascade model to simulate the influence diffusion. Upon the diffusion model, we explored the effects of different advance publicities for influence maximization.
{"title":"Influence Maximization within Period under Different Advance Publicity","authors":"Chih-Hua Tai, Ya-Wen Teng","doi":"10.1109/ICASI57738.2023.10179532","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179532","url":null,"abstract":"As social network platforms are becoming more and more popular, many activities spread event information through these platforms so that the number of participants during the event can reach the maximum. Note that as time goes by, the event information may be forgotten and repeatedly seen by the users. So the participation intention of a user for the event will change between active and inactive. Therefore, we formulate an influence maximization problem concerning the previous publicity and volatility of user behaviors given a specific period, and propose a fluctuation-aware independent-cascade model to simulate the influence diffusion. Upon the diffusion model, we explored the effects of different advance publicities for influence maximization.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124026716","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：