Proceedings of the 8th International Conference on Computing and Artificial Intelligence最新文献

Sparse DARTS with Various Recovery Algorithms 各种恢复算法的稀疏dart

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532225

Yanqing Hu, Qing Ye, Huan-Shuan Fu, Jiancheng Lv

{"title":"Sparse DARTS with Various Recovery Algorithms","authors":"Yanqing Hu, Qing Ye, Huan-Shuan Fu, Jiancheng Lv","doi":"10.1145/3532213.3532225","DOIUrl":"https://doi.org/10.1145/3532213.3532225","url":null,"abstract":"Designing an efficient neural architecture search method is an open and challenging problem over the last few years. A typical and well-performed strategy is gradient-based methods (i.e., Differentiable Architecture Search (DARTS)), which mainly searches the target sparse child graph from a trainable dense super graph. However, during the searching phrase, training the dense super graph usually requires excessively computational resources. Besides, the training based on a dense graph is excessively inefficient, and the memory consumption is prohibitively high. To alleviate this shortcoming, recently Iterative Shrinkage Thresholding Algorithm (ISTA), a sparse coding recovery algorithm, has been applied to DARTS, which directly optimizes the compressed representation of the super graph, and saves the memory and time consumption. Indeed, there are several such kinds of sparse coding recovery algorithms, and ISTA is not the best one in terms of recovery efficiency and effectiveness. To investigate the impact of different sparse coding recovery algorithm on performance in DARTS and provide some insights. Firstly, we design several sparse DARTS based on different sparse coding recovery algorithms (i.e., LISTA, CoD, and Lars). Then a series of controlled experiments on selected algorithms are conducted. The accuracy, search time and other indicators of the model are collected and compared. Sufficient theoretical analysis and experimental exploration reveal that the different compression algorithms show different characteristics on the sparse DARTS. Specifically, Lars-NAS tends to choose the operation with fewer parameters, while Cod-NAS is the simplest of the four recovery algorithms, and its consuming time is very short, but the CoD-NAS model is unstable. Particularly, LISTA-NAS achieves the accurate results with stable recovery time. Thus, it can be seen that all compression algorithms are available to utilized according to different environments and requirements.","PeriodicalId":333199,"journal":{"name":"Proceedings of the 8th International Conference on Computing and Artificial Intelligence","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123111000","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: 1

A Robust Recognition Method for Automotive Manufacturing Based on Deep Neural Networks 基于深度神经网络的汽车制造鲁棒识别方法

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532220

Li Li, Yanni Wang

引用次数: 0

Identifying review spam with an unsupervised approach based on topic abuse 基于主题滥用的无监督方法识别评论垃圾邮件

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532265

Jiandun Li, N. Li, Liu Yang, Pengpeng Zhang

引用次数: 0

Single Image 3D Reconstruction Based on Attention Mechanism and Graph Convolution Network 基于注意机制和图卷积网络的单幅图像三维重建

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532315

Wei Gao, Liyang Yu, Yuanyuan Du, Songfeng Lu

引用次数: 1

Lightweight Text Matching Method with Rich Features 具有丰富特征的轻量级文本匹配方法

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532262

Changhua Ji, Zhang Tao, Jiayi Mao, Li Zhang

{"title":"Lightweight Text Matching Method with Rich Features","authors":"Changhua Ji, Zhang Tao, Jiayi Mao, Li Zhang","doi":"10.1145/3532213.3532262","DOIUrl":"https://doi.org/10.1145/3532213.3532262","url":null,"abstract":"Text matching is one of the research hotspots in Natural Language Processing (NLP). The study of text matching is of great practical importance for applications such as text de-duplication, web retrieval, and question answering systems. A lightweight text matching method with rich features is proposed for the problem of large number of model parameters and low efficiency of text matching tasks in natural language processing. The whole model architecture is based on Siamese neural networks with shared parameters. Furthermore, the method utilizes an improved residual Network and attention mechanism for the extraction and alignment of vector representations. Only three key features for alignment operations are retained. In addition, an averaging operation is added to the fusion layer to provide vector representations with rich information for the prediction layer. Experimental results on the paraphrase identification dataset and two natural language inference datasets show that the proposed approach not only effectively reduces the number of parameters compared with existing models but also ensures good text matching performance. Experiments demonstrate that this method can be used in general text matching tasks.","PeriodicalId":333199,"journal":{"name":"Proceedings of the 8th International Conference on Computing and Artificial Intelligence","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114301136","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

Design of Speech Recognition Robot 语音识别机器人的设计

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532246

D. Wu, Baoqing Deng, Xiutian Zhuang

引用次数: 0

Infrared and visible image fusion based on edge-preserving filter and weighted least square optimization 基于边缘保持滤波和加权最小二乘优化的红外与可见光图像融合

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532338

Di Kang, Xin Zheng, Qiang Wu, Jinling Cui

{"title":"Infrared and visible image fusion based on edge-preserving filter and weighted least square optimization","authors":"Di Kang, Xin Zheng, Qiang Wu, Jinling Cui","doi":"10.1145/3532213.3532338","DOIUrl":"https://doi.org/10.1145/3532213.3532338","url":null,"abstract":"Infrared (IR) and visible (VI) image fusion play an important role in improving ability to scene perception and target detection, however, due to different imaging principles, significant feature differences of images make it very difficult to extract and integrate feature information effectively, especially in complex scenes where the target feature has different scales and contrast. Therefore, this paper proposes an image fusion method based on scale-aware edge-preserving filter and weighted least square optimization, aiming to extract features at different scales more accurately. First, we designed a hybrid feature decomposition method based on the scale-aware structure-preserving filter and Gaussian filter. The proposed method separated source images into region, structure, and texture layers, and thus achieved a finer-scale division than traditional multiscale decomposition methods. Then, according to the characteristics of infrared and visible images in the region layer and texture layer, the weighted least squares optimization framework is used combing with visual saliency map and scale-aware mechanism respectively, to obtain better visual expression effect. Experimental results indicated that the proposed method could achieve better subjective and objective results than current state-of-the-art methods.","PeriodicalId":333199,"journal":{"name":"Proceedings of the 8th International Conference on Computing and Artificial Intelligence","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124550692","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

Using Approximately Coupled Tensor Factorization to Model Changing User Preferences for Movie Recommendations 使用近似耦合张量分解来模拟电影推荐中用户偏好的变化

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532257

Yang Leng, Dehong Qiu

引用次数: 0

An Effective Approach for Large Ontology Matching Using Multi-objective Grasshopper Algorithm 基于多目标Grasshopper算法的大型本体匹配方法

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532230

Zhaoming Lv

{"title":"An Effective Approach for Large Ontology Matching Using Multi-objective Grasshopper Algorithm","authors":"Zhaoming Lv","doi":"10.1145/3532213.3532230","DOIUrl":"https://doi.org/10.1145/3532213.3532230","url":null,"abstract":"Although the population-based metaheuristic ontology matching approaches have achieved excellent results on small scale matching tasks, such methods do not solve the large ontology matching problem. In addition, the common practice of the ontology matching community is to divide the large ontology into many small fragments. Although the divide-and-conquer strategy is feasible to reduce time and space complexity, it can easily change the original structure of the ontology, resulting in reduced quality. Further, if the produced fragments are smaller, the number of fragments will increase, which introduces new time and space complexity. In this paper, an effective approach for large ontology matching using multi-objective metaheuristic Grasshopper algorithm is proposed, called GOLOM. In this approach, an ontology pruning technique is proposed to reduce time and space complexity while maintaining the original structure. An effective background knowledge is built to assist the basic matcher in a proper way. In order to demonstrate the performance of GOLOM, three large ontology matching tasks were conducted. Experimental results show that GOLOM significantly reduces time and memory complexity compared to partition-based. In terms of alignment quality, GOLOM outperforms all the state-of-the-art systems.","PeriodicalId":333199,"journal":{"name":"Proceedings of the 8th International Conference on Computing and Artificial Intelligence","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126435005","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

Analysis of public sentiment tendency in COVID-19 period based on GA-BiLSTM 基于GA-BiLSTM的新冠疫情时期民情趋势分析

Proceedings of the 8th International Conference on Computing and Artificial Intelligence Pub Date : 2022-03-18 DOI: 10.1145/3532213.3532242

Yingxue Tao, Zhenhua He, Yuhang Zhang

引用次数: 0