Pub Date : 2023-03-01DOI: 10.1109/prmvia58252.2023.00043
Ying Zhang, Di Peng, Gong Meng, Qian Zhao, Tiantian Li
This paper studies the fault diagnosis model of aeronautical multivariate heterogeneous input data. Because of the gyroscope’s powerful nonlinear mapping capabilities, it is a natural fit for modeling failure detection, this article combined with a variety of aviation gyro input data with fault monitoring methods, a model simulation method for multivariate heterogeneous input data in different states is proposed, which are one-dimensional and multi-dimensional data fault diagnosis in the standby state of the aircraft, and multi-sensor fault detection in the flight state or stationary state, which can effectively meet the needs of managing the fault diagnosis of multi-heterogeneous input of aviation.
{"title":"Simulation of Fault Diagnosis Model for Managing Aeronautical Multivariate Heterogeneous Inputs","authors":"Ying Zhang, Di Peng, Gong Meng, Qian Zhao, Tiantian Li","doi":"10.1109/prmvia58252.2023.00043","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00043","url":null,"abstract":"This paper studies the fault diagnosis model of aeronautical multivariate heterogeneous input data. Because of the gyroscope’s powerful nonlinear mapping capabilities, it is a natural fit for modeling failure detection, this article combined with a variety of aviation gyro input data with fault monitoring methods, a model simulation method for multivariate heterogeneous input data in different states is proposed, which are one-dimensional and multi-dimensional data fault diagnosis in the standby state of the aircraft, and multi-sensor fault detection in the flight state or stationary state, which can effectively meet the needs of managing the fault diagnosis of multi-heterogeneous input of aviation.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127704518","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 : 2023-03-01DOI: 10.1109/prmvia58252.2023.00055
Yunhao Pan, Chenhong Sui, Haipeng Wang, Hao Liu, Guobin Yang, Ao Wang, Q. Gong
Deep salient object detection has experienced noticeable progress. Unfortunately, most existing methods focus on clean samples regardless of the noise disturbance induced by human or natural factors. This results in the detection performance being extremely vulnerable to small perturbations. To this end, this paper proposes robust salient object detection via adversarial training (ATSOD). In specific, we introduce the classical DSS algorithm and inject it into an adversarial training framework favoring salient object detection. This ensures that, apart from clean samples, adversarial examples involving tiny disturbances are also explored for model training. Comparative experiments are conducted on five popular benchmarks. Experimental results show that despite the slight performance degradation for natural examples, there is a significant performance improvement for adversarial examples.
{"title":"Robust Salient Object Detection via Adversarial Training","authors":"Yunhao Pan, Chenhong Sui, Haipeng Wang, Hao Liu, Guobin Yang, Ao Wang, Q. Gong","doi":"10.1109/prmvia58252.2023.00055","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00055","url":null,"abstract":"Deep salient object detection has experienced noticeable progress. Unfortunately, most existing methods focus on clean samples regardless of the noise disturbance induced by human or natural factors. This results in the detection performance being extremely vulnerable to small perturbations. To this end, this paper proposes robust salient object detection via adversarial training (ATSOD). In specific, we introduce the classical DSS algorithm and inject it into an adversarial training framework favoring salient object detection. This ensures that, apart from clean samples, adversarial examples involving tiny disturbances are also explored for model training. Comparative experiments are conducted on five popular benchmarks. Experimental results show that despite the slight performance degradation for natural examples, there is a significant performance improvement for adversarial examples.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127765750","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 : 2023-03-01DOI: 10.1109/prmvia58252.2023.00013
Yi Zuo, Yun Zhou, Shengzong Liu, Yupeng Liu
Social tagging information to describe characteristics. Recent systems introduce tagging user preferences and item work shows that the recommendation accuracy can be remarkably promoted when tag information is handled properly. However, other performance indicators of recommendations, such as diversity and novelty, are also of great importance in practice. Thus, we propose a two-stage tag-aware multi-objective framework for providing accurate and diversity recommendations. Specifically, we formulate a tag-based recommendation algorithm via deep learning to generate accurate items and abstract effective tag-based potential features for users and items. According to these features, two conflicting objectives are designed to estimate the recommendation accuracy and diversity, respectively. By optimizing these two objectives simultaneously, the designed multi-objective recommendation model can pro-vide a set of recommendation lists for each user. Comparative experiments verify that the proposed model is promising to generate improved recommendations in terms of accuracy and diversity.
{"title":"A Tag-aware Recommendation Algorithm Based on Deep Learning and Multi-objective Optimization","authors":"Yi Zuo, Yun Zhou, Shengzong Liu, Yupeng Liu","doi":"10.1109/prmvia58252.2023.00013","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00013","url":null,"abstract":"Social tagging information to describe characteristics. Recent systems introduce tagging user preferences and item work shows that the recommendation accuracy can be remarkably promoted when tag information is handled properly. However, other performance indicators of recommendations, such as diversity and novelty, are also of great importance in practice. Thus, we propose a two-stage tag-aware multi-objective framework for providing accurate and diversity recommendations. Specifically, we formulate a tag-based recommendation algorithm via deep learning to generate accurate items and abstract effective tag-based potential features for users and items. According to these features, two conflicting objectives are designed to estimate the recommendation accuracy and diversity, respectively. By optimizing these two objectives simultaneously, the designed multi-objective recommendation model can pro-vide a set of recommendation lists for each user. Comparative experiments verify that the proposed model is promising to generate improved recommendations in terms of accuracy and diversity.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130341459","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 : 2023-03-01DOI: 10.1109/prmvia58252.2023.00025
Linghao Cai
Ceramic art has been passed down to the present day. It reflects the wisdom of ancient craftsmen and artists and is a treasure of Chinese culture. People with different cultural backgrounds and different artistic cultivation, their works are rooted in tradition and bold innovation. They are not only traditional and highly skilled skills, but also the essence of national culture. In recent years, with the rapid development of digital media technology, it has provided new development opportunities for the protection and inheritance of intangible cultural heritage. With the intervention of digital technology, the extension of ceramic design is also constantly extending, which requires ceramic designers to continuously expand their knowledge and combine multiple professional subject theories to enrich the connotation of their works. An excellent pottery work is not only a superficial artistic expression, but also a deep cultural heritage and innovative digital media performance. This article will discuss that in the digital age, the digital form of ceramics is obtained through three-dimensional scanning, and the ceramic art is digitally protected; and then the ceramic art is displayed through virtual technology and holographic imaging technology.
{"title":"Digital Protection and Virtual Display Technology of Ceramic Art","authors":"Linghao Cai","doi":"10.1109/prmvia58252.2023.00025","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00025","url":null,"abstract":"Ceramic art has been passed down to the present day. It reflects the wisdom of ancient craftsmen and artists and is a treasure of Chinese culture. People with different cultural backgrounds and different artistic cultivation, their works are rooted in tradition and bold innovation. They are not only traditional and highly skilled skills, but also the essence of national culture. In recent years, with the rapid development of digital media technology, it has provided new development opportunities for the protection and inheritance of intangible cultural heritage. With the intervention of digital technology, the extension of ceramic design is also constantly extending, which requires ceramic designers to continuously expand their knowledge and combine multiple professional subject theories to enrich the connotation of their works. An excellent pottery work is not only a superficial artistic expression, but also a deep cultural heritage and innovative digital media performance. This article will discuss that in the digital age, the digital form of ceramics is obtained through three-dimensional scanning, and the ceramic art is digitally protected; and then the ceramic art is displayed through virtual technology and holographic imaging technology.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122810186","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 : 2023-03-01DOI: 10.1109/PRMVIA58252.2023.00035
Xinjia Xie, Shun Gai, Han Long
Rumor detection on online social network (OSN) aims to help people retrieve reliable information and prevent public panic when emergencies occur suddenly. However, it is a waste of human efforts to detect rumors from the rapid growth of large-scale datasets. Due to the development of artificial intelligence, many architectures and frameworks are proposed to provide solutions for this issue. The first proposed traditional feature related methods are time-consuming and heavily depend on well-designed features, which calls for novel methods to detect rumors more efficiently. Thus deep neural networks related methods are successively born, and recent research on propagation related methods has captured much attention of both academia and industry. However, there lacks a systematic and global survey in the field of modern rumor detection. In this paper, we introduce rumors and OSN, and then present a comprehensive study of rumor detection methods on OSN, classifying them according to their search approaches and providing a comparison of the selected works. Finally, this survey deliver unique views on key challenges and several future research directions of rumor detection on OSN, such as multi-task learning, multi-modal detection and developing standard datasets and benchmarks. This work is supported by the Department of System Science, College of Liberal Arts and Sciences in National University of Defense Technology.
网络社交网络(online social network, OSN)的谣言检测旨在帮助人们在突发事件发生时找回可靠的信息，防止公众恐慌。然而，从快速增长的大规模数据集中检测谣言是浪费人力。由于人工智能的发展，人们提出了许多架构和框架来解决这个问题。首先提出的传统特征相关方法耗时长，并且严重依赖于精心设计的特征，这就需要新的方法来更有效地检测谣言。因此，与深度神经网络相关的方法相继诞生，而近年来对传播相关方法的研究也引起了学术界和工业界的广泛关注。然而，在现代谣言检测领域缺乏系统的、全面的研究。在本文中，我们介绍了谣言和OSN，然后对OSN上的谣言检测方法进行了全面的研究，根据它们的搜索方式对它们进行了分类，并对所选作品进行了比较。最后，本研究对基于OSN的谣言检测的关键挑战和未来的几个研究方向，如多任务学习、多模式检测和开发标准数据集和基准提出了独特的看法。本研究得到国防科技大学文理学院系统科学系的支持。
{"title":"Modern Techniques for Rumor Detection from the Perspective of Natural Language Processing","authors":"Xinjia Xie, Shun Gai, Han Long","doi":"10.1109/PRMVIA58252.2023.00035","DOIUrl":"https://doi.org/10.1109/PRMVIA58252.2023.00035","url":null,"abstract":"Rumor detection on online social network (OSN) aims to help people retrieve reliable information and prevent public panic when emergencies occur suddenly. However, it is a waste of human efforts to detect rumors from the rapid growth of large-scale datasets. Due to the development of artificial intelligence, many architectures and frameworks are proposed to provide solutions for this issue. The first proposed traditional feature related methods are time-consuming and heavily depend on well-designed features, which calls for novel methods to detect rumors more efficiently. Thus deep neural networks related methods are successively born, and recent research on propagation related methods has captured much attention of both academia and industry. However, there lacks a systematic and global survey in the field of modern rumor detection. In this paper, we introduce rumors and OSN, and then present a comprehensive study of rumor detection methods on OSN, classifying them according to their search approaches and providing a comparison of the selected works. Finally, this survey deliver unique views on key challenges and several future research directions of rumor detection on OSN, such as multi-task learning, multi-modal detection and developing standard datasets and benchmarks. This work is supported by the Department of System Science, College of Liberal Arts and Sciences in National University of Defense Technology.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128990580","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 : 2023-03-01DOI: 10.1109/prmvia58252.2023.00038
Tian Ni, Haoji Hu
Knowledge distillation is an efficient method in neural network compression, which transfers the knowledge from a high-capacity teacher network to a low-capacity student network. Previous approaches follow the ‘one teacher and one student’ paradigm, which neglects the possibility that interaction of multiple students could boost the distillation performance. In this paper, we propose a novel approach by simultaneously training multiple instances of a student model. By adding the similarity and diversity losses into the baseline knowledge distillation and adaptively adjusting the proportion of these losses according to accuracy changes of multiple student instances, we build a distillation system to make students collaborate and compete with each other, which improves system robustness and performance. Experiments show superior performance of the proposed method over existing offline and online distillation schemes on datasets with various scales.
{"title":"Knowledge Distillation by Multiple Student Instance Interaction","authors":"Tian Ni, Haoji Hu","doi":"10.1109/prmvia58252.2023.00038","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00038","url":null,"abstract":"Knowledge distillation is an efficient method in neural network compression, which transfers the knowledge from a high-capacity teacher network to a low-capacity student network. Previous approaches follow the ‘one teacher and one student’ paradigm, which neglects the possibility that interaction of multiple students could boost the distillation performance. In this paper, we propose a novel approach by simultaneously training multiple instances of a student model. By adding the similarity and diversity losses into the baseline knowledge distillation and adaptively adjusting the proportion of these losses according to accuracy changes of multiple student instances, we build a distillation system to make students collaborate and compete with each other, which improves system robustness and performance. Experiments show superior performance of the proposed method over existing offline and online distillation schemes on datasets with various scales.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126419042","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 : 2023-03-01DOI: 10.1109/PRMVIA58252.2023.00028
Juanqin Yan, Quan Zhou, Ya Xiao, Bin Pan
CART algorithm is a tree structure used for classification rules in the form of decision tree from a group of unordered and irregular cases. Compared with other classification methods, it has the advantage that a busy large amount of data can is classified yen fully, and then valuable potential information can be found. The method is simple and intuitive, with fast classification speed and high accuracy, which is suitable for large-scale data processing. Moreover, the algorithm process is easy to understand and can though express the importance of attributes praying attributes. The significant sensitivity and unpredictability of house price make it difficult to construct its forecasting model. In this paper, through an example of house price, the influencing factors of house price are deeply analyzed and the existing research results are systematically sorted out, and the decision tree CART detailed is used to build a molybdenum metal price algorithm model and forecast the actual price. By comparing and analyzing the results by using Not principles, the average absolute error is 4.03%, and the accuracy rate of foreforetrend forecasting trend can reach 94.8%, which shows that the algorithm is not only not intuitive and intuitive, but also reasonable and reliable.
{"title":"A Comparative Study Of CART Algorithm For Forecasting","authors":"Juanqin Yan, Quan Zhou, Ya Xiao, Bin Pan","doi":"10.1109/PRMVIA58252.2023.00028","DOIUrl":"https://doi.org/10.1109/PRMVIA58252.2023.00028","url":null,"abstract":"CART algorithm is a tree structure used for classification rules in the form of decision tree from a group of unordered and irregular cases. Compared with other classification methods, it has the advantage that a busy large amount of data can is classified yen fully, and then valuable potential information can be found. The method is simple and intuitive, with fast classification speed and high accuracy, which is suitable for large-scale data processing. Moreover, the algorithm process is easy to understand and can though express the importance of attributes praying attributes. The significant sensitivity and unpredictability of house price make it difficult to construct its forecasting model. In this paper, through an example of house price, the influencing factors of house price are deeply analyzed and the existing research results are systematically sorted out, and the decision tree CART detailed is used to build a molybdenum metal price algorithm model and forecast the actual price. By comparing and analyzing the results by using Not principles, the average absolute error is 4.03%, and the accuracy rate of foreforetrend forecasting trend can reach 94.8%, which shows that the algorithm is not only not intuitive and intuitive, but also reasonable and reliable.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125737698","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 : 2023-03-01DOI: 10.1109/PRMVIA58252.2023.00050
Jiayi Sun, Mingbo Zhao
Sentiment analysis is a crucial task in the research of natural language handling. Traditional machine learning approaches frequently employ bag-of-word representations that do not capture complex linguistic phenomena. The recursive autoencoder (RAE) method can availably learn the vector space representation of phrases, which is superior to other sentiment prediction methods on commonly used data sets. However, during the learning process, extensive label data is often required to label each node. In addition, RAE uses greedy strategies to merge adjacent words, it is difficult to capture long-distance and deeper semantic information. We put forward a semi-supervised approach that combines the SenticNet lexicon to train the recursive autoencoder for calculating the sentiment orientation of each node, and incorporates an attention mechanism to capture the contextual relationship between the words in a sentence. Experiments prove that the model proposed in this paper outperforms RAE and other models.
{"title":"Attention-Based Recursive Autoencoder For Sentence-Level Sentiment Classification","authors":"Jiayi Sun, Mingbo Zhao","doi":"10.1109/PRMVIA58252.2023.00050","DOIUrl":"https://doi.org/10.1109/PRMVIA58252.2023.00050","url":null,"abstract":"Sentiment analysis is a crucial task in the research of natural language handling. Traditional machine learning approaches frequently employ bag-of-word representations that do not capture complex linguistic phenomena. The recursive autoencoder (RAE) method can availably learn the vector space representation of phrases, which is superior to other sentiment prediction methods on commonly used data sets. However, during the learning process, extensive label data is often required to label each node. In addition, RAE uses greedy strategies to merge adjacent words, it is difficult to capture long-distance and deeper semantic information. We put forward a semi-supervised approach that combines the SenticNet lexicon to train the recursive autoencoder for calculating the sentiment orientation of each node, and incorporates an attention mechanism to capture the contextual relationship between the words in a sentence. Experiments prove that the model proposed in this paper outperforms RAE and other models.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125896713","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 : 2023-03-01DOI: 10.1109/prmvia58252.2023.00056
Xia Yu, YU Peng, Le Xia, Yuanrong He
In this paper, we monitor the surface deformation of Helan Mountains by using the DInSAR (Differential Interferometric Synthetic Aperture Radar) technology and Sentinel-1 SAR data from December 2019 to December 2021. The surface deformation of the Helan Mountain National Natural Reserve with a study area extending to 1935 km2 are observed. The findings indicate that the surface of Helan Mountain Reserve is rising in the east and sinking in the west, with no obvious increasing tendency in the north or south of Helan Mountain. Additionally, the vertical deformation map created by D-InSAR processing is used to monitor two monitoring cycles with significant deformations in June 2020 and December 2021. Furthermore, Helan Mountain has experienced two earthquakes with magnitudes of 3 or greater, according to the differential interference technique. An important decision-making basis for disaster prevention and mitigation can be provided by the deformation data of the ground surface obtained by the InSAR technology.
{"title":"Surface deformation monitoring based on DINSAR technique","authors":"Xia Yu, YU Peng, Le Xia, Yuanrong He","doi":"10.1109/prmvia58252.2023.00056","DOIUrl":"https://doi.org/10.1109/prmvia58252.2023.00056","url":null,"abstract":"In this paper, we monitor the surface deformation of Helan Mountains by using the DInSAR (Differential Interferometric Synthetic Aperture Radar) technology and Sentinel-1 SAR data from December 2019 to December 2021. The surface deformation of the Helan Mountain National Natural Reserve with a study area extending to 1935 km2 are observed. The findings indicate that the surface of Helan Mountain Reserve is rising in the east and sinking in the west, with no obvious increasing tendency in the north or south of Helan Mountain. Additionally, the vertical deformation map created by D-InSAR processing is used to monitor two monitoring cycles with significant deformations in June 2020 and December 2021. Furthermore, Helan Mountain has experienced two earthquakes with magnitudes of 3 or greater, according to the differential interference technique. An important decision-making basis for disaster prevention and mitigation can be provided by the deformation data of the ground surface obtained by the InSAR technology.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114063899","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 : 2001-05-02DOI: 10.1109/PRMVIA58252.2023.00037
Chao Hu, Weibin Qiu, Weijie Wu, Liqiang Zhu
Video anomaly detection (VAD) detects target objects such as people and vehicles to discover abnormal events in videos. There are abundant spatio-temporal context information in different objects of videos. Most existing methods pay more attention to temporal context than spatial context in VAD. The spatial context information represents the relationship between the detection target and surrounding targets. Anomaly detection makes a lot of sense. To this end, a video anomaly detection algorithm based on target spatio-temporal context fusion is proposed. Firstly, the target in the video frame is extracted through the target detection network to reduce background interference. Then the optical flow map of two adjacent frames is calculated. Motion features are used multiple targets in the video frame to construct spatial context simultaneously, re-encoding the target appearance and motion features, and finally reconstructing the above features through the spatiotemporal dual-stream network, and using the reconstruction error to represent the abnormal score. The algorithm achieves frame-level AUCs of 98.5% on UCSDped2 and 86.3% on Avenue datasets. On UCSDped2 dataset, the spatio-temporal dual-stream network improves frames by 5.1% and 0.3%, respectively, compared to the temporal and spatial stream networks. After using spatial context encoding, the frame-level AUC is enhanced by 1%, which verifies the method’s effectiveness.
{"title":"Spatio-Temporal-based Context Fusion for Video Anomaly Detection","authors":"Chao Hu, Weibin Qiu, Weijie Wu, Liqiang Zhu","doi":"10.1109/PRMVIA58252.2023.00037","DOIUrl":"https://doi.org/10.1109/PRMVIA58252.2023.00037","url":null,"abstract":"Video anomaly detection (VAD) detects target objects such as people and vehicles to discover abnormal events in videos. There are abundant spatio-temporal context information in different objects of videos. Most existing methods pay more attention to temporal context than spatial context in VAD. The spatial context information represents the relationship between the detection target and surrounding targets. Anomaly detection makes a lot of sense. To this end, a video anomaly detection algorithm based on target spatio-temporal context fusion is proposed. Firstly, the target in the video frame is extracted through the target detection network to reduce background interference. Then the optical flow map of two adjacent frames is calculated. Motion features are used multiple targets in the video frame to construct spatial context simultaneously, re-encoding the target appearance and motion features, and finally reconstructing the above features through the spatiotemporal dual-stream network, and using the reconstruction error to represent the abnormal score. The algorithm achieves frame-level AUCs of 98.5% on UCSDped2 and 86.3% on Avenue datasets. On UCSDped2 dataset, the spatio-temporal dual-stream network improves frames by 5.1% and 0.3%, respectively, compared to the temporal and spatial stream networks. After using spatial context encoding, the frame-level AUC is enhanced by 1%, which verifies the method’s effectiveness.","PeriodicalId":221346,"journal":{"name":"2023 International Conference on Pattern Recognition, Machine Vision and Intelligent Algorithms (PRMVIA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2001-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126900951","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：