Guanping Liang, Junyan Chen, Junlin Xie, Zhangjia Deng, Yiwen Cui
{"title":"Classification of fine-grained species of marine organisms based on multi-scale fusion","authors":"Guanping Liang, Junyan Chen, Junlin Xie, Zhangjia Deng, Yiwen Cui","doi":"10.1145/3485314.3485321","DOIUrl":"https://doi.org/10.1145/3485314.3485321","url":null,"abstract":"Due to the limited shooting conditions of marine life, image data is often difficult to obtain; The shooting is greatly affected by factors such as light, visibility, etc, which leads to shortcomings such as color cast and blurring of the image. Some species have very little difference in characteristics, making it difficult to distinguish visually. It is for the above reasons that the classification of marine organisms is facing great difficulties. Based on the characteristics of marine biological image data sets, this paper has carried out research on classification schemes of marine biological fine-grained species. Initially, the EfficientNet-b5 single model is used for training. During the training process, tricks such as Cutmix, Autoaugment, label smoothing, and CBAM modules are added, and the final accuracy is as high as 96.59%. Since features of different scales have different semantic information, the solution improves the single model, uses EfficientNet-b3 and EfficientNet-b6 for multi-scale fusion, and embeds the CBAM mechanism into different scales to obtain richer features information. Finally, experiments were conducted on the existing ocean image data set, and the accuracy rate reached 97.82%, which confirmed the feasibility and effectiveness of the scheme.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116249046","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}
{"title":"Performance Analysis of Star Network for Distributed Unmanned Combat Platforms based on LTE","authors":"Hongbim Luo, Shengbo Hu, Fang Ming, Man-Qin Zhu","doi":"10.1145/3485314.3485629","DOIUrl":"https://doi.org/10.1145/3485314.3485629","url":null,"abstract":"Distributed unmanned combat is a new combat style shaping future information warfare, for which network communication is crucial. This study constructed the star network topology of LTE-based distributed unmanned combat platform against the background of using distributed unmanned combat vehicles in future combat. The network performances of the distributed unmanned combat vehicle and LTE wireless base station communication under the star topology are simulated and analyzed using the discrete-event network simulator NS3. The results show that the star structure is adaptable in the LTE network; the uplink and downlink performance indicators of the distributed unmanned combat vehicles and LTE wireless base station reach a high level; and the data can be transmitted effectively.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132219735","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}
Aidong Xu, Xuechun Wang, Yunan Zhang, Tao Wu, Xingping Xian
{"title":"Adversarial Attacks on Deep Neural Networks for Time Series Prediction","authors":"Aidong Xu, Xuechun Wang, Yunan Zhang, Tao Wu, Xingping Xian","doi":"10.1145/3485314.3485316","DOIUrl":"https://doi.org/10.1145/3485314.3485316","url":null,"abstract":"Time series data is widespread in real-world scenarios. To recover and infer missing information in practical domains, such as stock price monitoring, electricity load forecasting, traffic flows analysis, climate trend prediction, etc., the problem of time series prediction has been widely studied as a classical research topic in data mining. Over the past decade, deep learning architectures are introduced as a vital part of the next generation of time series prediction models. However, recent studies showed that deep learning models are vulnerable to adversarial attacks. In this paper, we study the adversarial attacks on the time series prediction models prospectively. We propose an attack strategy to generate adversarial samples by adding imperceptible perturbed data to the original time series with the goal of reducing the accuracy of time series prediction models. Specifically, the perturbation-based adversarial example generation algorithm is proposed using gradient information of time series prediction model. Moreover, adversarial examples should be imperceptible to humans. To address the challenge, we craft adversarial samples based on importance measuring to perturb the original data locally. We evaluate our attacks on state-of-the-art time series prediction models using three time series datasets. Our results demonstrate that our attacks can effectively evade the time series prediction models, and the adversarial attacks mechanisms can be used as robustness metric for constructing robust time series prediction models.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131763542","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}
{"title":"KPI anomaly detection method for Data Center AIOps based on GRU-GAN","authors":"Hang Su, Qian He, Biao Guo","doi":"10.1145/3485314.3485323","DOIUrl":"https://doi.org/10.1145/3485314.3485323","url":null,"abstract":"The system architecture and application services of the data center are becoming increasingly large. To ensure the stable operation of the systems and businesses carried by the data center, the operations engineer needs to collect and monitor the generating KPIs during the operation of the systems and services. Traditional KPI anomaly detection methods are faced with the challenges of the huge amount of KPIs and constantly changing data characteristics, which are gradually no longer suitable for highly dynamic systems and services. With the popularity of artificial intelligence algorithms, machine learning and deep learning methods have also begun to be applied in operation and maintenance scenarios, that is the emergence of Artificial Intelligence for IT Operations (AIOps). KPI anomaly detection is the underlying core technology of AIOps. This paper proposes a hybrid model based on GRU-GAN (GGAN) for KPI anomaly detection in data center AIOps. The Gated Recurrent Unit (GRU) network is selected as the generator and discriminator of Generative adversarial network (GAN) in this model, which get the time correlation and data distribution of KPI through the adversarial training between the generator and the discriminator to make use of the reconstruction ability of the generator and the discriminant ability of the discriminator at the same time. At the anomaly detection stage, the anomaly score is formed by integrating reconstruction difference and discrimination loss to complete the anomaly detection task. Experimental results show that the proposed method can more accurately capture the variable data characteristics of KPI compared with the traditional KPI anomaly detection method and the general unsupervised method, as well as achieve better performance in the KPI anomaly detection task.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114939717","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}
{"title":"A scalable ideal progressive visual cryptography scheme","authors":"Lingfu Wang, Jing Wang, Weijia Huang","doi":"10.1145/3485314.3485320","DOIUrl":"https://doi.org/10.1145/3485314.3485320","url":null,"abstract":"Visual cryptography is a kind of image-based secret sharing scheme to share secret image based on the human visual system (HVS). For instance, a ()-visual cryptography scheme (VCS) encrypts a secret image into shares and distributes them to different participants, the secret image can be recovered by at least shares. However, most of the existing VCSs are designed for binary images with limited flexibility and scalability. In this paper, a scalable ideal progressive visual cryptography scheme (PVCS) is proposed by scalable basis matrices and random Gaussian noise, which can process non-binarized images directly without pixel expansion. Our proposed basis matrix expansion method greatly improves the flexibility of the scheme and thus the efficiency is improved. In addition, since the visual effect of the VCS is standardized by the HVS and lacks an accurate quantitative index, we measure our scheme by using the peak signal-to-noise ratio (PSNR) and the structural similarity (SSIM), we also propose a method to derive the noise volume according to SSIM. Furthermore, sufficient experiments show that the proposed scheme is visually efficient and superior.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129531362","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}
{"title":"Fusion of infrared and visible light images for object detection based on CNN","authors":"Dong Liu, Huihua Yang, Yuying Shao","doi":"10.1145/3485314.3485327","DOIUrl":"https://doi.org/10.1145/3485314.3485327","url":null,"abstract":"In this paper, using the advantages of convolutional neural networks in feature extraction, we design feature extraction modules. According to the different fusion positions of the extracted features in the network, several feature fusion schemes are designed to perform global feature fusion on infrared images and visible light images. A detection model based on the regression of the center point and the target scale is proposed. The fused feature map is used as the input of the object detection module to obtain the detection in the current image field of view. It effectively solves the problem of poor performance of visible light images in object detection caused by light intensity and weather conditions.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116616618","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}
{"title":"Multi-granularity context semantic fusion model for Chinese event detection","authors":"Xiaokao Tan, Guofeng Deng, Xiangjun Hu","doi":"10.1145/3485314.3485322","DOIUrl":"https://doi.org/10.1145/3485314.3485322","url":null,"abstract":"Event detection is a key task in the field of information extraction, which is widely used in knowledge mapping, intelligent question answering, reading comprehension and other fields. Event detection model based on deep learning commomly regards event detection as a word classification task, but in Chinese, a language without natural separators, the error of word segmentation will lead to the mismatch between word and event trigger. In addition, the polysemy of Chinese words can make a word ambiguous in different contexts. In this paper, we propose a Chinese event detection model based on multi granularity context semantic fusion. Firstly, the semantic information in different word segmentation results is obtained through the Character-Word fusion gate mechanism to solve the problem of mismatch between Chinese words and event triggers. Then, the Character-Sentence fusion gate is designed to learn the semantic information of the whole sequence, and the self-attention mechanism is used to integrate the contextual semantic information to eliminate the ambiguity of Chinese words. Experiments on ACE2005 dataset show that our method can achieve better experimental results than the current mainstream Chinese event detection model.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116283956","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}
Jianhong Feng, Yuwei Zhou, Zhaohong Yang, Hongyu Wang
{"title":"Design and implementation of special Data Encyclopedia system","authors":"Jianhong Feng, Yuwei Zhou, Zhaohong Yang, Hongyu Wang","doi":"10.1145/3485314.3485330","DOIUrl":"https://doi.org/10.1145/3485314.3485330","url":null,"abstract":"Abstract:The traditional data construction in the military field mainly focuses on the vertical business field, realizing the digitization of the military business and failing to meet the application requirements of horizontal data fusion.This paper realizes the design of a special data encyclopedia system in the military field, and uses advanced big data processing technologies such as web crawler, knowledge spectrum, and full-text retrieval to realize the horizontal integration of data in the military field, mine the value of data in the military field, and improve the practicability and timeliness of data.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130416418","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}
{"title":"A multilayer linear regression approach to find the coefficients of two dimensional values from the live weight and the manual body measurement process of a cow.","authors":"M. Masudur Rahman, Sk. Md. Masudul Ahsan","doi":"10.1145/3485314.3485329","DOIUrl":"https://doi.org/10.1145/3485314.3485329","url":null,"abstract":"In the Livestock sector, cattle's weight measurement is always a challenge. Manual methods are used to get an idea of the weight and body mass of the cow. The factors that influence the variation of real live weight and the flesh weight from the manual body measurement of a cow depend on various parameters of the cattle. That may be the age, height, breed, cow types like bull, ox, milking cow, calf, etc. In this study, we tried to identify important parameters that play a more influential role in changing the outcome. In the study, we determined the weight of the cow in two ways. One is live weight by scale (using weight measuring machine), the other is weight by length and girth (using live weight calculation equation). A machine learning technique named the multilayer linear regression method finally used to determine the coefficient of the most influential parameters. More than eight hundred data from different areas and varieties of cows have been collected to ensure the accuracy of the measurement.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"166 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115320061","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}
{"title":"Convolutional Neural Networks for Artificial Marbling Beef Classification","authors":"Grace Panitchakorn, Y. Limpiyakorn","doi":"10.1145/3485314.3485332","DOIUrl":"https://doi.org/10.1145/3485314.3485332","url":null,"abstract":"Nowadays, image processing is widely used as one of the quality control techniques applied in the food and neutral industrial sectors. One of the problem domains deep learning excels is image classification. From a deep learning perspective, a variety of image classification problems could be quickly solved through transfer learning. This paper thus presents a transfer learning solution for the artificial marbling beef classification problem. The transfer learning from two selected pretrained models: VGG16 and InceptionV3 was carried out to construct deep convolutional neural networks models for binary classification if the image is artificial marbling beef. The preliminary result showed that the CNN+InceptionV3 outperformed the other two models which are the CNN trained by scratch and the CNN+VGG16. The intensely, beautifully marbled fat results in rich flavor and increased price of the cuts. The proposed approach of image classification is promising and ,more or less, the classifier would benefit the buyers for detecting imposter marbling beef with mark-up price.","PeriodicalId":321724,"journal":{"name":"2021 10th International Conference on Internet Computing for Science and Engineering","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126269986","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}