Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things最新文献

A Remote Fatigue Driving Detection System for Ship Supervision based on Physiological Response Features 基于生理反应特征的船舶监管疲劳驾驶远程检测系统

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603813

Jinming Tong, Wei Cheng, Chen Li, Xuming Wang, Guan-Chun Chen

{"title":"A Remote Fatigue Driving Detection System for Ship Supervision based on Physiological Response Features","authors":"Jinming Tong, Wei Cheng, Chen Li, Xuming Wang, Guan-Chun Chen","doi":"10.1145/3603781.3603813","DOIUrl":"https://doi.org/10.1145/3603781.3603813","url":null,"abstract":"Fatigue driving is one of the main influential factors causing maritime accidents, traditional physiological signal detection method has disadvantages such as poor stability and practicability, it has great individual differences and always interferes with the driver operation. This paper proposes a remote fatigue driving detection system based on physiological response features. By fusing different physiological response features such as head posture and eye closure, a fatigue detection model is constructed. As a supplement to single EAR detection for reducing the missed retrieval of eye closure behavior, Single Shot Multi Box Detector is applied to improve the accuracy and robustness of the system. The PERCLOS value is approximately solved by the number of frames with eye closure, and the abnormal head posture angle and the P80 standard have been used to evaluate the fatigue state. Experimental result shows that the detection accuracy has reached 96.9548%, it could meet the demand of ship supervision for driving behavior and fatigue detection which has prosperous application value in seafarers' training and maritime management fields.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121834959","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}

引用次数: 0

Flood hazard assessment model based on ST-UNet 基于ST-UNet的洪水灾害评价模型

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603914

Haoran Ran

引用次数: 0

A Reliable Blockchain Computation Offloading Scheme for Internet of Things 一种可靠的物联网区块链计算卸载方案

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603895

Chuanjia Yao, Youquan Xian, Chunpei Li, Dongcheng Li, Boyi Wang, Ying Zhao

{"title":"A Reliable Blockchain Computation Offloading Scheme for Internet of Things","authors":"Chuanjia Yao, Youquan Xian, Chunpei Li, Dongcheng Li, Boyi Wang, Ying Zhao","doi":"10.1145/3603781.3603895","DOIUrl":"https://doi.org/10.1145/3603781.3603895","url":null,"abstract":"At present, the integration of blockchain and Internet of Things (IoT) has become a popular concept. Blockchain can provide IoT with a trusted data-sharing and computing environment. Yet the on-chain computing efficiency of blockchain is still limited by the underlying consensus algorithms, but blockchain can offload computation-intensive tasks to off-chain IoT nodes to improve its performance. However, most of the existing computation offloading schemes are incapable to guarantee the reliability of off-chain computing in untrustworthy IoT environments since participants can collude. Therefore, this work aims at improving the reliability of computation offloading. We introduce a voting-based game into our system workflow to enhance the confidence of computation results. Meanwhile, an anonymous node selection algorithm, which depends on a priority assignment method based on Verifiable Random Function (VRF), is designed to improve the quality of selected nodes and avoid the collusion between participants. The extensive results shows that the proposed scheme can achieve competitive performance in terms of fairness, accuracy, and robustness.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126948518","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}

引用次数: 0

A joint improved DETR network face recognition algorithm based on TSM module and DANN. 基于TSM模块和DANN的联合改进DETR网络人脸识别算法。

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603916

Zihe Ye

{"title":"A joint improved DETR network face recognition algorithm based on TSM module and DANN.","authors":"Zihe Ye","doi":"10.1145/3603781.3603916","DOIUrl":"https://doi.org/10.1145/3603781.3603916","url":null,"abstract":"As one of the most mature application fields of artificial intelligence, face recognition system has been widely used in production and life. But at the same time as large-scale commercialization, face recognition technology also faces more challenges. How to further improve the accuracy of face recognition function and improve the defense ability of the recognition system against face against samples is an important research direction of algorithms. At present, the algorithm mostly focuses on the picture features of a single portrait, ignores the details difference in the time domain of the fake video, and has the problems of weak generalization ability and overfitting of the model. In this paper, an improved DETR network is proposed, which uses the TSM module to perform time domain displacement on the extracted video features and averages the time features of pooled learning samples to better distinguish dynamic adversarial sample instances. At the same time, DANN is introduced as a systematic classification and discrimination network, which uses the domain classifier to domain discrimination of the feature space and uses the adversarial loss function to update the feature extractor and domain classifier parameters. Actual tests show that the recognition accuracy of the network on the FaceForensics dataset is improved by an average of 4%-7% compared with the improvement period, the error rate is less than 7%, and the model recognition speed index is not higher 400ms, which proves that the model has high accuracy rate and good real-time solving ability.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122883475","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}

引用次数: 0

LanYUAN, a GPT large model using Curriculum Learning and Sparse Attention 基于课程学习和稀疏注意的GPT大模型LanYUAN

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603827

Gonghai Zhou, Yuhong Zhang, Rizhen Hu, Yang Zhang

{"title":"LanYUAN, a GPT large model using Curriculum Learning and Sparse Attention","authors":"Gonghai Zhou, Yuhong Zhang, Rizhen Hu, Yang Zhang","doi":"10.1145/3603781.3603827","DOIUrl":"https://doi.org/10.1145/3603781.3603827","url":null,"abstract":"In 2021, the Inspur AI Research Institute introduced the AI Megatron Model Yuan-1.0, a massive Chinese language AI model containing 245.7 billion parameters. This model surpassed OpenAI's GPT-3, making it the world's largest Chinese NLP model. Although the model was pre-trained using Nvidia's Megatron framework with model parallelism, data parallelism, and pipelining optimizations, there is still room for improvement in terms of training time, cost, and convergence. To achieve better performance, this paper investigates the impacts of batch size and learning rate on model training time and accuracy to balance model performance. We replaced the pipelining optimization with the more efficient DeepSpeed framework, and combined DeepSpeed's ZeRO-based data parallelism with Nvidia's Megatron-LM model parallelism to achieve higher performance on Nvidia GPU clusters with high-bandwidth interconnects. Additionally, we used a curriculum learning-based method and four types of sparse attention as a new optimization approaches. The results showed that the training time was reduced by 20% and the throughput increased by 20% compared to the 47 billion parameters Yuan-1.0 model. Approximately, the optimized model achieved performance improvement in downstream tasks with the same training data.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"265 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128772040","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}

引用次数: 0

Genetic algorithm and dynamic planning based airline shipment optimization system 基于遗传算法和动态规划的航空货运优化系统

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603892

Zi-Fang Li, Siyuan Cheng, Shirui Tang

{"title":"Genetic algorithm and dynamic planning based airline shipment optimization system","authors":"Zi-Fang Li, Siyuan Cheng, Shirui Tang","doi":"10.1145/3603781.3603892","DOIUrl":"https://doi.org/10.1145/3603781.3603892","url":null,"abstract":"Aiming at the difficult NP problem of box selection and crating in the airline shipment process, this paper constructs an airline shipment optimization system based on genetic algorithm and dynamic planning. It can be solved under multiple constraints to get the cargo loading scheme with the highest utilization of crating space and the highest economic efficiency. In this paper, the cargo is firstly organized according to the average value of the first fifty cycles, and the optimization system is designed to leave as little gap as possible. Considering the container dimension constraint, a container selection optimization model which integrates the container volume, dead weight and space utilization is established, and the optimal container type corresponding to each cargo is solved by genetic algorithm. After that, a trinomial tree structure model is established for the containers, and a spatial partitioning algorithm is designed to solve the packing scheme for the cargoes and further obtain the required number of containers. The optimal system of airline shipment considering the economic efficiency and the reliability of changing cargo source is studied by the combined optimization algorithm and the time series algorithm respectively. The reliability is determined to be 95%, and the conclusions obtained prove that the optimization system proposed in this paper can complete the shipment task allocation work quickly and efficiently.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130590125","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}

引用次数: 0

An Efficient and Reliable I/O Mapping Protocol for Industrial Cyber-Physical Systems 一种高效可靠的工业信息物理系统I/O映射协议

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603883

Songxi Liu, Liang Shi, Zhiwei Liu

{"title":"An Efficient and Reliable I/O Mapping Protocol for Industrial Cyber-Physical Systems","authors":"Songxi Liu, Liang Shi, Zhiwei Liu","doi":"10.1145/3603781.3603883","DOIUrl":"https://doi.org/10.1145/3603781.3603883","url":null,"abstract":"With the advancement of Industry 4.0, industrial cyber-physical systems (ICPS) are expected to realize the digitalization and intelligence in smart factories, and lots of sensors and actuators will be deployed and networked together. However, limited bandwith resource and massive data volume make it difficult to meet the real-time and reliability requirement in the industrial network. To address these problems, an efficient and reliable Input/Output (I/O) mapping protocol is proposed to provide the quality of service. First, we establish an I/O mapping communication model for smart manufacturing and automation scenarios, in which devices communicate with each other by simply writing and reading the output and input area. In order to ensure the reliability of transmission, we introduce an optimized hand-shake mechanism with a small hand-shake overhead. Finally, to realize the real-time transmission with coexistance of large-scale sensor data and control data, a bandwidth reservation method is designed to alleviate interference from each other. Extensive evaluation results show that our proposed protocol performs well in terms of packet-loss rate and transmission delay, and the intuitive I/O communication interface can improve the efficiency of software development.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126415036","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}

引用次数: 0

Multi-scale 2.5D Integrated Model for False Positive Reduction in Abdominal Lymph Node Detection 腹部淋巴结检测中假阳性降低的多尺度2.5D集成模型

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603807

Changquan Lu, Kun Yu, Xukun Zhang, Wenxin Hu

{"title":"Multi-scale 2.5D Integrated Model for False Positive Reduction in Abdominal Lymph Node Detection","authors":"Changquan Lu, Kun Yu, Xukun Zhang, Wenxin Hu","doi":"10.1145/3603781.3603807","DOIUrl":"https://doi.org/10.1145/3603781.3603807","url":null,"abstract":"Accurate lymph node detection based on CT images is an important pre-step in Lymphadenectomy, which is crucial for the subsequent treatment of patients. However, many interfering factors such as organs, tissues, and blood vessels in CT images lead a large number of false positive objects with the existing detection methods, which hinders the clinical application. In this study, we propose an effective and fast 2.5D framework algorithm for false positive reduction (FPR). Specifically, considering the globular shape of the lymph nodes, the axial, coronal and sagittal slices from CT data of lymph nodes are used as 2.5D inputs to the network to mine the three-dimensional spatial information of lymph nodes. In addition, multi-scale inputs are used to address the challenges posed by changes in the volume of the lymph nodes, and use stacking learning to fuse the results of models at various scales. Compared to 3D networks, it greatly decreases computational costs and running time. Based on the public dataset, we add 20 additional cases of CT data to construct a new dataset CTLymph for FPR. The proposed method achieves an AUC of 0.941, reducing the number of false positive lymph nodes from 25 to 4. The result shows that our proposed model achieves superior performance and outperforms several state-of-the-art methods.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125978473","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}

引用次数: 0

Mask Multi-Head Attention with Partition Network for Vehicle Re-Identification 基于分区网络的车辆再识别方法

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603852

Yang Liu, Chen Kong, Yue-Ji Li, P. Zhang, Han Yu

{"title":"Mask Multi-Head Attention with Partition Network for Vehicle Re-Identification","authors":"Yang Liu, Chen Kong, Yue-Ji Li, P. Zhang, Han Yu","doi":"10.1145/3603781.3603852","DOIUrl":"https://doi.org/10.1145/3603781.3603852","url":null,"abstract":"Vehicle Re-identification aims to match a specific vehicle image across different places or cameras based on the similarity among vehicles. vehicle re-id remains confronted with two severe challenges, small inter-class variability caused by a similar vehicle with a similar type and color, and dramatic intra-class variability caused by the variation of view. More recently, methods are proposed to improve performance by using additional metadata such as critical points and orientation, which all require expensive annotations. Therefore, we introduce attention mechanism to solve these two problems without considering extra annotation. In this paper, we propose a novel mask multi-head attention with partition network (MMAPN). To discover subtle differences between two similar vehicles, we propose a partition unit to discover more local detail. To extract features that are robust to both tremendous intra-class differences and subtle inter-class variability, we propose a mask multi-head attention block to extract potential features. Extensive experimental evaluations show our approach achieved state-of-the-art performance.","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134107300","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}

引用次数: 0

Probing the influence maximization-cost minimization problem in social networks by using a multi-objective discrete differential evolution optimization 利用多目标离散微分进化优化方法探讨社交网络中影响最大化-成本最小化问题

Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things Pub Date : 2023-05-26 DOI: 10.1145/3603781.3603886

Jianxin Tang, Li Zhang, Pengli Lu, Jimao Lan

{"title":"Probing the influence maximization-cost minimization problem in social networks by using a multi-objective discrete differential evolution optimization","authors":"Jianxin Tang, Li Zhang, Pengli Lu, Jimao Lan","doi":"10.1145/3603781.3603886","DOIUrl":"https://doi.org/10.1145/3603781.3603886","url":null,"abstract":"Influence maximization is to extract k most influential individuals that can maximize the spreading coverage of the promoted information in social networks. The existing majority of efforts merely focus on how to maximize the coverage of the promotion. Whereas the disposable budget acts as a significant factor needed to be considered in practical scenarios. In this paper, we take into account both the influence maximization and cost minimization simultaneously in the perspective of variable cost for the activation of different candidate node with the size of targeted seed set fixed, and formulate the problem as a multi-objective optimization. A multi-objective discrete differential evolution optimization (MODDE) with mutation, crossover and selection operators specifically adaptable to the topological network structure is proposed. For the non-uniform cost setting for each node in the influence spreading process, a novel functional metric is designed to measure the cost of igniting each candidate node. The non-dominated solutions derived from MODDE can better balance the coverage of influence and budget costs, thus providing decision makers with more choices. Extensive experiments and statistic tests on real-world networks are performed to estimate the proposed method, and the results demonstrate the outperformance of the MODDE over the state-of-the-art methods. Additional Keywords and Phrases: Social network analysis, Influence maximization, Cost minimization, Multi-objective optimization, Differential evolution algorithm, Pareto optimal","PeriodicalId":391180,"journal":{"name":"Proceedings of the 2023 4th International Conference on Computing, Networks and Internet of Things","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133216604","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}

引用次数: 0