Pub Date : 2023-04-21DOI: 10.1109/ICASI57738.2023.10179533
Jueun Jeong, Han-Gyul Jeong, Han-joon Kim
This paper presents BAMTGAN, a novel data augmentation technique that addresses the class imbalance problem and prevents mode collapse by utilizing a modified DCGAN model and a new similarity loss to generate diverse and realistic tabular data. BAMTGAN encodes each column to produce a feature map for each record, which is then converted back to its original tabular form an intermediate image format. Experimental results demonstrate that BAMTGAN provides a more substantial improvement in developing high-quality predictive models than existing augmentation methods. Github: https://github.com/uos-dmlab/Structured-Data-Augmentation.git
{"title":"BAMTGAN: A Balanced Augmentation Technique for Tabular Data","authors":"Jueun Jeong, Han-Gyul Jeong, Han-joon Kim","doi":"10.1109/ICASI57738.2023.10179533","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179533","url":null,"abstract":"This paper presents BAMTGAN, a novel data augmentation technique that addresses the class imbalance problem and prevents mode collapse by utilizing a modified DCGAN model and a new similarity loss to generate diverse and realistic tabular data. BAMTGAN encodes each column to produce a feature map for each record, which is then converted back to its original tabular form an intermediate image format. Experimental results demonstrate that BAMTGAN provides a more substantial improvement in developing high-quality predictive models than existing augmentation methods. Github: https://github.com/uos-dmlab/Structured-Data-Augmentation.git","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"159 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":"122616743","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/ICASI57738.2023.10179517
Yi-Yuan Chen, Kuo-Hsun Hsu, Andrew Weian Hou
The advancement of containerization and various integration tools has greatly reduced the development cost of microservice architecture. Under this premise, many companies have begun to migrate their legacy systems (monolithic applications) to microservice architectures in order to adapt to market demands. Instead of migrating legacy systems to microservice architecture from scratch, we provide another approach of splitting an existing monolithic system into microservices. In this paper, we use Go language’s AST as a basis to split the original project into multiple independent components and design a method for analyzing dependencies, classifying them, and producing independently runnable microservices.
{"title":"MAT: Automating Go monolithic applications transform into microservices through dependency analysis and AST","authors":"Yi-Yuan Chen, Kuo-Hsun Hsu, Andrew Weian Hou","doi":"10.1109/ICASI57738.2023.10179517","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179517","url":null,"abstract":"The advancement of containerization and various integration tools has greatly reduced the development cost of microservice architecture. Under this premise, many companies have begun to migrate their legacy systems (monolithic applications) to microservice architectures in order to adapt to market demands. Instead of migrating legacy systems to microservice architecture from scratch, we provide another approach of splitting an existing monolithic system into microservices. In this paper, we use Go language’s AST as a basis to split the original project into multiple independent components and design a method for analyzing dependencies, classifying them, and producing independently runnable microservices.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"230 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":"122734377","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/ICASI57738.2023.10179528
N. Shih, Yun-Ting Tasi
This study aimed to document and to interact with 3D geoscience in Keelung, Taiwan, by reconstructing the urban fabric using unmanned aerial vehicle (UAV), smartphone light detection and ranging (LiDAR), geographic information system (GIS), augmented reality (AR), and rapid prototyping (RP). We found that the highly fluctuated skyline concluded the intensity of development in history. The applications of modeling tools and cloud database enabled the interaction with physical and virtual context using smartphone in iOS® and Android® systems. The final result was able to present detailed 3D visual details of an old lane by cultural and property identity. This novel simulation presented an integrated process of heterogeneous 3D data from a small to a large scale in a sustainable regeneration of cultural identity.
{"title":"Urban Simulation in UAV, Smartphone, GIS, AR, and RP","authors":"N. Shih, Yun-Ting Tasi","doi":"10.1109/ICASI57738.2023.10179528","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179528","url":null,"abstract":"This study aimed to document and to interact with 3D geoscience in Keelung, Taiwan, by reconstructing the urban fabric using unmanned aerial vehicle (UAV), smartphone light detection and ranging (LiDAR), geographic information system (GIS), augmented reality (AR), and rapid prototyping (RP). We found that the highly fluctuated skyline concluded the intensity of development in history. The applications of modeling tools and cloud database enabled the interaction with physical and virtual context using smartphone in iOS® and Android® systems. The final result was able to present detailed 3D visual details of an old lane by cultural and property identity. This novel simulation presented an integrated process of heterogeneous 3D data from a small to a large scale in a sustainable regeneration of cultural identity.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"55 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":"122872003","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/ICASI57738.2023.10179499
Muhammad Usman, Chun-Ling Lin
Fifteen novices completed a Zentangle task while their brain activity was measured with electroencephalography (EEG) and their concentration emotional state, stress, and anxiety levels were evaluated with questionnaires. The study found increased activity in theta, alpha, beta, and gamma ranges associated with relaxation, focus, and mental activity. Participants reported improved concentration, emotional state, and reduced stress and anxiety levels after completing the task, suggesting the potential of Zentangle to enhance cognitive and emotional well-being.
{"title":"Using electroencephalography signals to explore the influence of Zentangle on meditation and mindfulness","authors":"Muhammad Usman, Chun-Ling Lin","doi":"10.1109/ICASI57738.2023.10179499","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179499","url":null,"abstract":"Fifteen novices completed a Zentangle task while their brain activity was measured with electroencephalography (EEG) and their concentration emotional state, stress, and anxiety levels were evaluated with questionnaires. The study found increased activity in theta, alpha, beta, and gamma ranges associated with relaxation, focus, and mental activity. Participants reported improved concentration, emotional state, and reduced stress and anxiety levels after completing the task, suggesting the potential of Zentangle to enhance cognitive and emotional well-being.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"1 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":"125407764","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/ICASI57738.2023.10179554
Shuo-Han Chen, Zhen-Yang Guo
With the booming of artificial intelligence and the advance of embedded systems, smart surveillance systems can now deploy video analysis techniques, such as human detection and object classification, on surveillance cameras directly. Such an approach decentralizes the computation and allows elastic camera deployment without upgrading the centralized server. Nevertheless, taking human detection as an example, cameras at different locations could have very different numbers of people walking by at different times. This condition could make specific cameras run their detection tasks at the highest frame per second (FPS) but still lose information if FPS is lower than a certain value, while a small group of cameras has almost zero tasks, thus wasting valuable computation power. In addition, as the smart surveillance system grows, cameras deployed at different periods could have different specifications and computation capabilities; therefore, properly utilizing all the computation resources has become essential for smart surveillance systems. The aforementioned observations motivate this study to propose a novel computation souring method by following the concept of crowdsourcing to enable efficient resource sharing within smart surveillance systems and maximize the amount of retrieved information. Promising results have been demonstrated through a series of experiments with different smart cameras of different specifications.
{"title":"Computation Sourcing in Smart Surveillance Systems for Maximum Information Retrieval","authors":"Shuo-Han Chen, Zhen-Yang Guo","doi":"10.1109/ICASI57738.2023.10179554","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179554","url":null,"abstract":"With the booming of artificial intelligence and the advance of embedded systems, smart surveillance systems can now deploy video analysis techniques, such as human detection and object classification, on surveillance cameras directly. Such an approach decentralizes the computation and allows elastic camera deployment without upgrading the centralized server. Nevertheless, taking human detection as an example, cameras at different locations could have very different numbers of people walking by at different times. This condition could make specific cameras run their detection tasks at the highest frame per second (FPS) but still lose information if FPS is lower than a certain value, while a small group of cameras has almost zero tasks, thus wasting valuable computation power. In addition, as the smart surveillance system grows, cameras deployed at different periods could have different specifications and computation capabilities; therefore, properly utilizing all the computation resources has become essential for smart surveillance systems. The aforementioned observations motivate this study to propose a novel computation souring method by following the concept of crowdsourcing to enable efficient resource sharing within smart surveillance systems and maximize the amount of retrieved information. Promising results have been demonstrated through a series of experiments with different smart cameras of different specifications.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"12 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":"121210589","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}
Quantized deep learning models are suitable for the embedded devices with limited computation resource. For computation-intensive neural network operators such as convolution, heterogeneous platforms with a set of processing units of different types become common in the embedded devices. These embedded devices usually operate on fixed-point calculations; moreover, they rely on customized kernel functions to deploy deep learning models. In this paper, a flow of deploying pre-quantized deep learning models on heterogeneous platforms using TVM is presented. We propose an optimization to convert quantization parameters. To leverage customized kernel functions, we propose the operator flow recognition. To demonstrate our flow, we utilize embARC Machine Learning Inference (embARC MLI), an open-source software library targeted for low-power applications. A set of pre-quantized deep learning models are deployed on a heterogeneous platform comprising x86 and embARC MLI. Experimental results show that for each model, the accuracy obtained from the heterogeneous platform is much the same as the one obtained from an x86 platform.
{"title":"Deploying Pre-Quantized Deep Learning Models on Heterogeneous Platforms with Operator Flow Recognition and Quantization Parameter Optimization","authors":"Kuen-Wey Lin, Yan-Ying Li, Kuan Wang, Ming-Chih Tung","doi":"10.1109/ICASI57738.2023.10179562","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179562","url":null,"abstract":"Quantized deep learning models are suitable for the embedded devices with limited computation resource. For computation-intensive neural network operators such as convolution, heterogeneous platforms with a set of processing units of different types become common in the embedded devices. These embedded devices usually operate on fixed-point calculations; moreover, they rely on customized kernel functions to deploy deep learning models. In this paper, a flow of deploying pre-quantized deep learning models on heterogeneous platforms using TVM is presented. We propose an optimization to convert quantization parameters. To leverage customized kernel functions, we propose the operator flow recognition. To demonstrate our flow, we utilize embARC Machine Learning Inference (embARC MLI), an open-source software library targeted for low-power applications. A set of pre-quantized deep learning models are deployed on a heterogeneous platform comprising x86 and embARC MLI. Experimental results show that for each model, the accuracy obtained from the heterogeneous platform is much the same as the one obtained from an x86 platform.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"106 5 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":"127458103","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/ICASI57738.2023.10179590
Yen-Liang Pan, W. Liao, Zi-Yang Wu, Jia-Cing Jhang, Fu-Rong Gu
This paper focuses on the use of Petri nets (PNs) to model a General Dynamics F-16 Fighting Falcon (F-16) missile launching system (MLS). The advances in digital computers and system-integration techniques have made the digitization of fight control computer systems possible. The fourth generation of fighters has used electronic digital signals to launch missiles. The most significant advantage of the digital MLS lies in its ability to cast without depending on the mechanical connecting rod. In other words, digital MLS will not break down due to mechanical failure. As a result, fighter pilots can better focus on their missions. A stable MLS is essential for launching missiles smoothly. Therefore, how to design one stable and easy-to-use MLS is important. In this paper, the authors tried to model and develop a stable MLS using the much easier-to-use Petri nets. One real-world digital MLS is used to verify the model. The experimental results using HPsim reveal that the designed model is correct and practical. Most importantly, the proposed model is much easier to use than the existing literature.
{"title":"Modeling and Enhancement of General Dynamics F-16 Fighting Falcon Missile Launching System Based on Petri Nets","authors":"Yen-Liang Pan, W. Liao, Zi-Yang Wu, Jia-Cing Jhang, Fu-Rong Gu","doi":"10.1109/ICASI57738.2023.10179590","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179590","url":null,"abstract":"This paper focuses on the use of Petri nets (PNs) to model a General Dynamics F-16 Fighting Falcon (F-16) missile launching system (MLS). The advances in digital computers and system-integration techniques have made the digitization of fight control computer systems possible. The fourth generation of fighters has used electronic digital signals to launch missiles. The most significant advantage of the digital MLS lies in its ability to cast without depending on the mechanical connecting rod. In other words, digital MLS will not break down due to mechanical failure. As a result, fighter pilots can better focus on their missions. A stable MLS is essential for launching missiles smoothly. Therefore, how to design one stable and easy-to-use MLS is important. In this paper, the authors tried to model and develop a stable MLS using the much easier-to-use Petri nets. One real-world digital MLS is used to verify the model. The experimental results using HPsim reveal that the designed model is correct and practical. Most importantly, the proposed model is much easier to use than the existing literature.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"1 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":"130271488","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/ICASI57738.2023.10179538
Huai-Kuei Wu, Yu-Chun Peng, Yen-Ming Chen, Yuji Wang
As society progresses, there are many different types of buildings compared to the past, including high-rise buildings, townhouses, and row houses. With the increasing complexity of building structures, evacuation guidance facilities have gradually improved. However, many people overlook their own home evacuation guidance. After all, ordinary homes cannot have “emergency exits” signs like public areas for the sake of aesthetics, resulting in huge losses. In 2021 alone, the financial losses were as high as NTD 356,205,000, and 192 people died as a result of fires. The number of building fires is highest for buildings between 1 to 5 floors, with a total of 4,734 incidents. In Taiwan, these buildings are mostly old apartments, which are much simpler in structure compared to community buildings. As long as a person can reach a position near the balcony, they can be rescued. However, they often do not have the opportunity to reach the balcony due to the severe environment of thick smoke and fire. Firefighters also face difficulties in searching and rescuing due to the dense and high-temperature smoke in the fire.This creation integrates sockets, smoke detectors, gateways, and mobile application software, and includes an OLED display and buttons on the socket end. The socket, smoke detector, and gateway are connected and can transmit data to a cloud platform via the Internet. The data transmitted during normal operation includes power consumption, smoke detection, temperature, etc.The above-mentioned cloud platform not only receives data from the gateway but also accepts connections from the mobile application. The mobile application can upload pre-set indoor escape routes to the cloud platform, providing reference for evacuation instructions in case of a fire. In addition, the mobile application can also obtain the usage status of all indoor sockets and control the power supply of the sockets when necessary. When a fire occurs, the OLED display on the socket can show the evacuation route according to the current fire situation, helping residents to evacuate. When residents are trapped, they can use the button on the socket to call for help. The rescuer can use the mobile application to locate the person in need of rescue, facilitating the rescue process.
{"title":"Development of Fire Escape System Integrated with Smart Socket","authors":"Huai-Kuei Wu, Yu-Chun Peng, Yen-Ming Chen, Yuji Wang","doi":"10.1109/ICASI57738.2023.10179538","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179538","url":null,"abstract":"As society progresses, there are many different types of buildings compared to the past, including high-rise buildings, townhouses, and row houses. With the increasing complexity of building structures, evacuation guidance facilities have gradually improved. However, many people overlook their own home evacuation guidance. After all, ordinary homes cannot have “emergency exits” signs like public areas for the sake of aesthetics, resulting in huge losses. In 2021 alone, the financial losses were as high as NTD 356,205,000, and 192 people died as a result of fires. The number of building fires is highest for buildings between 1 to 5 floors, with a total of 4,734 incidents. In Taiwan, these buildings are mostly old apartments, which are much simpler in structure compared to community buildings. As long as a person can reach a position near the balcony, they can be rescued. However, they often do not have the opportunity to reach the balcony due to the severe environment of thick smoke and fire. Firefighters also face difficulties in searching and rescuing due to the dense and high-temperature smoke in the fire.This creation integrates sockets, smoke detectors, gateways, and mobile application software, and includes an OLED display and buttons on the socket end. The socket, smoke detector, and gateway are connected and can transmit data to a cloud platform via the Internet. The data transmitted during normal operation includes power consumption, smoke detection, temperature, etc.The above-mentioned cloud platform not only receives data from the gateway but also accepts connections from the mobile application. The mobile application can upload pre-set indoor escape routes to the cloud platform, providing reference for evacuation instructions in case of a fire. In addition, the mobile application can also obtain the usage status of all indoor sockets and control the power supply of the sockets when necessary. When a fire occurs, the OLED display on the socket can show the evacuation route according to the current fire situation, helping residents to evacuate. When residents are trapped, they can use the button on the socket to call for help. The rescuer can use the mobile application to locate the person in need of rescue, facilitating the rescue process.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"61 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":"126020739","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/ICASI57738.2023.10179582
Yi-Cheng Hung, Zih-Yue Lin, L. Fu, Chia-Yen Lee
A Micro-Electro-Mechanical Systems based tungsten trioxide thin film micro-ozone sensor is prepared using physical vapor deposition techniques. Given a sufficiently high working temperature, the resistance of the tungsten trioxide film increases linearly with an increasing ozone concentration. The effects of variations in the working temperature on the resistance of the tungsten trioxide sensing layer are suppressed by a PID feedback control system implemented in LabVIEW. It is shown that a working temperature of 250°C results in the device’s maximum sensitivity (3.32 MΩ/ppm) and minimum response time (17 seconds).
{"title":"Design and Fabrication of Ozone Sensor based on WO3 thin film","authors":"Yi-Cheng Hung, Zih-Yue Lin, L. Fu, Chia-Yen Lee","doi":"10.1109/ICASI57738.2023.10179582","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179582","url":null,"abstract":"A Micro-Electro-Mechanical Systems based tungsten trioxide thin film micro-ozone sensor is prepared using physical vapor deposition techniques. Given a sufficiently high working temperature, the resistance of the tungsten trioxide film increases linearly with an increasing ozone concentration. The effects of variations in the working temperature on the resistance of the tungsten trioxide sensing layer are suppressed by a PID feedback control system implemented in LabVIEW. It is shown that a working temperature of 250°C results in the device’s maximum sensitivity (3.32 MΩ/ppm) and minimum response time (17 seconds).","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"1 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":"130270312","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/ICASI57738.2023.10179550
Shih-Li Lu, S. Miaou, Shyang-En Weng, Ying-Cheng Lin
Dehazing research is crucial to ensuring the safety of autonomous driving. To estimate the scattering coefficient of the scene, we use the point cloud produced by LiDAR. To acquire a more precise scene depth, we employ a stereo depth network. Finally, we dehaze the image using the transmission map of the atmospheric scattering model and the atmospheric light value. Experimental results show that the proposed dehazing method works better in object detection than previous dehazing methods.
{"title":"Using Stereo Depth Estimation Network and LiDAR-Assisted Camera for Dehazing","authors":"Shih-Li Lu, S. Miaou, Shyang-En Weng, Ying-Cheng Lin","doi":"10.1109/ICASI57738.2023.10179550","DOIUrl":"https://doi.org/10.1109/ICASI57738.2023.10179550","url":null,"abstract":"Dehazing research is crucial to ensuring the safety of autonomous driving. To estimate the scattering coefficient of the scene, we use the point cloud produced by LiDAR. To acquire a more precise scene depth, we employ a stereo depth network. Finally, we dehaze the image using the transmission map of the atmospheric scattering model and the atmospheric light value. Experimental results show that the proposed dehazing method works better in object detection than previous dehazing methods.","PeriodicalId":281254,"journal":{"name":"2023 9th International Conference on Applied System Innovation (ICASI)","volume":"14 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":"124584585","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：