Mobile edge caching (MEC) is a promising technology to alleviate traffic congestion in the network. Current studies explored deep reinforcement learning (DRL)-based MEC methods. These methods consider the dynamics of the request size to maximize the cache hit rate. However, they usually ignored the potential request relationships among contents. Two contents with a strong relationship are usually requested sequentially. Inspired by this assumption, this paper proposes a relationship-enhanced content caching algorithm using DRL, named RECC. Our RECC infers user preferences by mining the request relationships among contents. In this work, the relationships are modeled as request sequences, and the request features are learned by using graph embedding. These features will be used as input of state in our DRL-based algorithm. We utilize the Wolpertinger architecture to solve the limitation of large discrete action space. The simulation results indicate that our RECC outperformed the traditional cache policies and state-of-the-art DRL-based method in cache hit rate. Furthermore, the proposed RECC has advantages in long-term stability in the environment where content popularity changes dynamically, and also has a higher cache hit rate when handling the requests with number changes dynamically.
{"title":"RECC: A Relationship-Enhanced Content Caching Algorithm Using Deep Reinforcement Learning","authors":"Jiarui Ren, Haiyan Zhang, Xiaoping Zhou, Menghan Zhu","doi":"10.1109/ACAIT56212.2022.10137967","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137967","url":null,"abstract":"Mobile edge caching (MEC) is a promising technology to alleviate traffic congestion in the network. Current studies explored deep reinforcement learning (DRL)-based MEC methods. These methods consider the dynamics of the request size to maximize the cache hit rate. However, they usually ignored the potential request relationships among contents. Two contents with a strong relationship are usually requested sequentially. Inspired by this assumption, this paper proposes a relationship-enhanced content caching algorithm using DRL, named RECC. Our RECC infers user preferences by mining the request relationships among contents. In this work, the relationships are modeled as request sequences, and the request features are learned by using graph embedding. These features will be used as input of state in our DRL-based algorithm. We utilize the Wolpertinger architecture to solve the limitation of large discrete action space. The simulation results indicate that our RECC outperformed the traditional cache policies and state-of-the-art DRL-based method in cache hit rate. Furthermore, the proposed RECC has advantages in long-term stability in the environment where content popularity changes dynamically, and also has a higher cache hit rate when handling the requests with number changes dynamically.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132704626","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137976
Qiushi Hao, Jia Ren
Acoustic emission technology has a great advantage over existing nondestructive technologies for real-time inspection, which will significantly improve the efficiency of wheel/rail defect detection. However, wheel-rail rolling noise impedes the application of acoustic emission technology in on-line operation, especially in high-speed or heavy-load condition. The key problem lies in that current researches haven’t developed adequate knowledge of the noise, making it difficult to gain the defect signal under the strong noise. To study mechanical originations of the noise and reveal its intrinsic properties, a spectral analysis method is proposed based on a fractal description of rough surfaces. Power spectra of the surface and those of the noise, as well as the relation of their fractal dimensions, are investigated. Then, under the instruction of spectral distributions of microscopic mechanical behaviors, the noise originations and influence of the vehicle speed are determined. It is found that the noise is generated based on the surface topography, while sliding friction, particle behavior, and abrasive wear are the main mechanical sources. The sliding friction dominates among the three behaviors. The speed promotes all the behaviors and then enhances the power level, while its effects on the sliding friction is relatively severer. The work offers a theoretical basis and mechanical explanation for the noise, which provides further guidance for the real-time detection of defect signals.
{"title":"Determination of the Mechanical Origination of Wheel-Rail Rolling Noise Based on Spectrum Analysis","authors":"Qiushi Hao, Jia Ren","doi":"10.1109/ACAIT56212.2022.10137976","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137976","url":null,"abstract":"Acoustic emission technology has a great advantage over existing nondestructive technologies for real-time inspection, which will significantly improve the efficiency of wheel/rail defect detection. However, wheel-rail rolling noise impedes the application of acoustic emission technology in on-line operation, especially in high-speed or heavy-load condition. The key problem lies in that current researches haven’t developed adequate knowledge of the noise, making it difficult to gain the defect signal under the strong noise. To study mechanical originations of the noise and reveal its intrinsic properties, a spectral analysis method is proposed based on a fractal description of rough surfaces. Power spectra of the surface and those of the noise, as well as the relation of their fractal dimensions, are investigated. Then, under the instruction of spectral distributions of microscopic mechanical behaviors, the noise originations and influence of the vehicle speed are determined. It is found that the noise is generated based on the surface topography, while sliding friction, particle behavior, and abrasive wear are the main mechanical sources. The sliding friction dominates among the three behaviors. The speed promotes all the behaviors and then enhances the power level, while its effects on the sliding friction is relatively severer. The work offers a theoretical basis and mechanical explanation for the noise, which provides further guidance for the real-time detection of defect signals.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132738506","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137930
Pan Gao, Zhuhua Hu
Due to the diversity of human body posture, there are problems such as occlusion of key points, difference of target scale and background blur among people. Therefore, multi-human pose estimation is still a challenging task. The existing deep learning-based multi-body pose estimation methods are mainly divided into top-down and bottom-up, but most of them do not make full use of local features in the network. In this paper, convolutional block attention module(CBAM) and Focal L2 Loss were used to process the context information of convolutional neural network and consolidate local features. Specifically, we propose attention-containing stacked hourglass network (ASHN). ASHN is based on a stacked hourglass network, with the addition of a convolutional block attention module (CBAM) module to improve performance, combined with Focal L2 Loss in the model. Compared with the existing methods, our method achieves competitive performance, achieving 66.8% AP, 72.1% AP75 and 65.4% APM on COCO data sets.
{"title":"ASHN for Multi-Human Pose Estimation","authors":"Pan Gao, Zhuhua Hu","doi":"10.1109/ACAIT56212.2022.10137930","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137930","url":null,"abstract":"Due to the diversity of human body posture, there are problems such as occlusion of key points, difference of target scale and background blur among people. Therefore, multi-human pose estimation is still a challenging task. The existing deep learning-based multi-body pose estimation methods are mainly divided into top-down and bottom-up, but most of them do not make full use of local features in the network. In this paper, convolutional block attention module(CBAM) and Focal L2 Loss were used to process the context information of convolutional neural network and consolidate local features. Specifically, we propose attention-containing stacked hourglass network (ASHN). ASHN is based on a stacked hourglass network, with the addition of a convolutional block attention module (CBAM) module to improve performance, combined with Focal L2 Loss in the model. Compared with the existing methods, our method achieves competitive performance, achieving 66.8% AP, 72.1% AP75 and 65.4% APM on COCO data sets.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123182182","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137851
Y. Huang
In order to improve the accuracy of English education level evaluation, this paper puts forward a design method of associated estimation model of English education level based on artificial neural network. Establish a multiattribute decision-making constraint parameter model for the correlation assessment of English education level, and analyze the multi-attribute decision-making and quantitative characteristics of the correlation assessment of English education level combined with the multi-dimensional explanatory variable and control variable parameter identification methods. Combined with the artificial neural network modeling method, the feature clustering analysis of the English education level is carried out; the adaptive learning and training method of the artificial neural network is used to establish the attribute fusion set and the semantic ontology feature distribution set of the multi-attribute decision-making for the correlation evaluation of the English education level; using the artificial neural network The network output layer fusion control method realizes the optimization of the multiattribute decision-making process. The simulation results show that the method has a good effect on the intelligent decisionmaking of the correlation evaluation of English education level, and improves the accuracy of the evaluation results of English education level.
{"title":"Association Estimation of English Education Level Using Artificial Neural Network Algorithm","authors":"Y. Huang","doi":"10.1109/ACAIT56212.2022.10137851","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137851","url":null,"abstract":"In order to improve the accuracy of English education level evaluation, this paper puts forward a design method of associated estimation model of English education level based on artificial neural network. Establish a multiattribute decision-making constraint parameter model for the correlation assessment of English education level, and analyze the multi-attribute decision-making and quantitative characteristics of the correlation assessment of English education level combined with the multi-dimensional explanatory variable and control variable parameter identification methods. Combined with the artificial neural network modeling method, the feature clustering analysis of the English education level is carried out; the adaptive learning and training method of the artificial neural network is used to establish the attribute fusion set and the semantic ontology feature distribution set of the multi-attribute decision-making for the correlation evaluation of the English education level; using the artificial neural network The network output layer fusion control method realizes the optimization of the multiattribute decision-making process. The simulation results show that the method has a good effect on the intelligent decisionmaking of the correlation evaluation of English education level, and improves the accuracy of the evaluation results of English education level.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115123098","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137892
Siyu Wang, Dayong Dai
In order to further improve the temperature regulation performance of building phase change materials (PCM), on the basis of the traditional regulation hours, a multiobjective function which integrates the regulation hours and the economy of the envelope structure was proposed, and the objective function is solved by genetic algorithm (GA). The experimental results show that the optimal combination of each attribute of phase change material can be obtained by genetic algorithm, and the regulation hours and contribution rate obtained by genetic algorithm are more advantageous than PSO solution method. This shows that the room temperature of building phase change materials can be better solved through genetic algorithm, so as to achieve the purpose of joint improvement of temperature and economy, and has certain reference value.
{"title":"Multi-Objective Optimization of Room Temperature Regulation of Building Phase Change Materials Based on Genetic Algorithm","authors":"Siyu Wang, Dayong Dai","doi":"10.1109/ACAIT56212.2022.10137892","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137892","url":null,"abstract":"In order to further improve the temperature regulation performance of building phase change materials (PCM), on the basis of the traditional regulation hours, a multiobjective function which integrates the regulation hours and the economy of the envelope structure was proposed, and the objective function is solved by genetic algorithm (GA). The experimental results show that the optimal combination of each attribute of phase change material can be obtained by genetic algorithm, and the regulation hours and contribution rate obtained by genetic algorithm are more advantageous than PSO solution method. This shows that the room temperature of building phase change materials can be better solved through genetic algorithm, so as to achieve the purpose of joint improvement of temperature and economy, and has certain reference value.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"171 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122139475","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137971
Hui Ma, Zhuhua Hu, Yan Zheng
When consumers choose to buy clothing online, virtual try-on technology can provide them with a better shopping experience. The optimization of virtual try-on technology not only helps consumers to evaluate the selected clothing, but also can improve the profit for merchants. However, the traditional virtual try-on technology has problems such as high cost, image distortion, and deviation of clothing style. In order to solve the above problems, this paper proposes a virtual try-on model with enhanced feature representation capability. Through the improved residual block of Squeeze-and-Excitation Networks (SENet) and the style encoding module introduced by the Pyramid Squeeze Attention (PSA) module, our model enriches the content and style information, strengthens the representation ability of features, and the reconstructed image preserves the more details. Compared with related work, we improve the structural similarity measure by 1.1% and the Inception Score by 10.1%. It is demonstrated that our model can reconstruct more accurate and realistic images.
{"title":"A Virtual Try-on Model with Enhanced Feature Representation Capability","authors":"Hui Ma, Zhuhua Hu, Yan Zheng","doi":"10.1109/ACAIT56212.2022.10137971","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137971","url":null,"abstract":"When consumers choose to buy clothing online, virtual try-on technology can provide them with a better shopping experience. The optimization of virtual try-on technology not only helps consumers to evaluate the selected clothing, but also can improve the profit for merchants. However, the traditional virtual try-on technology has problems such as high cost, image distortion, and deviation of clothing style. In order to solve the above problems, this paper proposes a virtual try-on model with enhanced feature representation capability. Through the improved residual block of Squeeze-and-Excitation Networks (SENet) and the style encoding module introduced by the Pyramid Squeeze Attention (PSA) module, our model enriches the content and style information, strengthens the representation ability of features, and the reconstructed image preserves the more details. Compared with related work, we improve the structural similarity measure by 1.1% and the Inception Score by 10.1%. It is demonstrated that our model can reconstruct more accurate and realistic images.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"87 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132882673","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137839
Zhuohao Wu, Yanni Li, Danwen Ji, Dingming Wu, M. Shidujaman, Yuan Zhang, Chenfan Zhang
Artificial intelligence (AI) is trained with data, especially texts, numbers, images, videos and music on Internet. These data all together across time and space make a collective memory of the world. The latest large-scale AI models give people a chance to create out of a large pool of this collective memory, which they won’t be able to access before, and communicate with AI in both human natural language and the unique machine supported ways. As demonstrated and discussed in this paper, effective and efficient workflows can be built up for human and AI to co-create meaningful results based on both the collective memory of the world and the personalized ideas and tastes. This kind of human-AI co-creation has a new force with great potential, and expect a new strategy and philosophy to guide human-AI collaboration.
{"title":"Human-AI Co-Creation of Art Based on the Personalization of Collective Memory","authors":"Zhuohao Wu, Yanni Li, Danwen Ji, Dingming Wu, M. Shidujaman, Yuan Zhang, Chenfan Zhang","doi":"10.1109/ACAIT56212.2022.10137839","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137839","url":null,"abstract":"Artificial intelligence (AI) is trained with data, especially texts, numbers, images, videos and music on Internet. These data all together across time and space make a collective memory of the world. The latest large-scale AI models give people a chance to create out of a large pool of this collective memory, which they won’t be able to access before, and communicate with AI in both human natural language and the unique machine supported ways. As demonstrated and discussed in this paper, effective and efficient workflows can be built up for human and AI to co-create meaningful results based on both the collective memory of the world and the personalized ideas and tastes. This kind of human-AI co-creation has a new force with great potential, and expect a new strategy and philosophy to guide human-AI collaboration.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129035335","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137872
Yiping Rong, Han Su, Wenxin Zhang, Zhongyan Li
Fine-grained image classification is very difficult task and there is currently no effective machine learning method yet. Taking the butterfly images as an example, this paper comprehensively studies image classification methods for finegrained and complex images, focusing on improving the existing methods in the four aspects of image preprocessing, feature extraction, feature coding, and classifier design. An effective machine learning method for butterfly classification is established. In the aspect of image preprocessing, we firstly introduce the edge breakpoint connection method to make up for the defect that discontinuous edge can’t extract the target region effectively. In terms of feature extraction, nonseparable Shannon wavelet is used to combine with corner feature, blob feature, edge curvature feature and invariant moment feature selection to further improve the utilization of image information. In the aspect of feature encoding, most features in this paper are encoded by histogram, which increases the classification efficiency, and for histogram encoding features, an improved histogram intersection distance is proposed, which makes it more effective in KNN classifier. Finally, in the classifier design, the Bagging ensemble method is integrated into the parallel KNN classifier. Experiments show the effectiveness and robustness of the proposed method.
{"title":"Fine-Grained Complex Image Classification Method Based on Butterfly Images","authors":"Yiping Rong, Han Su, Wenxin Zhang, Zhongyan Li","doi":"10.1109/ACAIT56212.2022.10137872","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137872","url":null,"abstract":"Fine-grained image classification is very difficult task and there is currently no effective machine learning method yet. Taking the butterfly images as an example, this paper comprehensively studies image classification methods for finegrained and complex images, focusing on improving the existing methods in the four aspects of image preprocessing, feature extraction, feature coding, and classifier design. An effective machine learning method for butterfly classification is established. In the aspect of image preprocessing, we firstly introduce the edge breakpoint connection method to make up for the defect that discontinuous edge can’t extract the target region effectively. In terms of feature extraction, nonseparable Shannon wavelet is used to combine with corner feature, blob feature, edge curvature feature and invariant moment feature selection to further improve the utilization of image information. In the aspect of feature encoding, most features in this paper are encoded by histogram, which increases the classification efficiency, and for histogram encoding features, an improved histogram intersection distance is proposed, which makes it more effective in KNN classifier. Finally, in the classifier design, the Bagging ensemble method is integrated into the parallel KNN classifier. Experiments show the effectiveness and robustness of the proposed method.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"275 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129174271","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137997
Shalley Bakshi, Surbhi Sharma, R. Khanna
Optimization in energy-harvesting cognitive radio networks is accomplished by the amalgamation of metaheuristics with wireless networks. The design of an optimized energy-harvesting cognitive radio network (EHCRN) is challenging in the realm of wireless networks. This paper proposes a modified optimization technique rank-based multiobjective antlion optimization (RMOALO) based on antlions that finds an approximate solution to the optimization problem of sensing duration and energy consumption with throughput maximization. The search behavior of antlions is improved thus reaching an optimal solution while considering the constraints on collision and energy. The simulated results obtained in this paper show that the average throughput of the secondary wireless network gets maximized for an optimized sensing duration. The results also demonstrate the effect of spectrum sensing duration on the average harvested energy and average throughput for the energy-sufficient and energy-deficit regions.
{"title":"Performance Optimization in Energy Harvesting Cognitive Radio Networks a Shift Towards Metaheuristics","authors":"Shalley Bakshi, Surbhi Sharma, R. Khanna","doi":"10.1109/ACAIT56212.2022.10137997","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137997","url":null,"abstract":"Optimization in energy-harvesting cognitive radio networks is accomplished by the amalgamation of metaheuristics with wireless networks. The design of an optimized energy-harvesting cognitive radio network (EHCRN) is challenging in the realm of wireless networks. This paper proposes a modified optimization technique rank-based multiobjective antlion optimization (RMOALO) based on antlions that finds an approximate solution to the optimization problem of sensing duration and energy consumption with throughput maximization. The search behavior of antlions is improved thus reaching an optimal solution while considering the constraints on collision and energy. The simulated results obtained in this paper show that the average throughput of the secondary wireless network gets maximized for an optimized sensing duration. The results also demonstrate the effect of spectrum sensing duration on the average harvested energy and average throughput for the energy-sufficient and energy-deficit regions.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125641963","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 : 2022-12-09DOI: 10.1109/ACAIT56212.2022.10137983
Meixia Li
This paper mainly uses the comparative analysis method and the case analysis method to explain the explainable artificial intelligence (XAI) in music. The use of XAI in music is real-time interactive, and the more explainable and transparent, the more accurate it is. Interpretability can occur in or after modeling, and deep learning provides theoretical support for XAI. Professor Nick Bryan-Kinns' team made a breakthrough through experimental research on “XAI+Music” with the difficulties currently being explained by artificial intelligence, XAI technology has been directly applied to the creative AI model, for “XAI+Music” development and innovation provide references and ideas.
{"title":"The Use of Explainable Artificial Intelligence in Music—Take Professor Nick Bryan-Kinns’ “XAI+Music” Research as a Perspective","authors":"Meixia Li","doi":"10.1109/ACAIT56212.2022.10137983","DOIUrl":"https://doi.org/10.1109/ACAIT56212.2022.10137983","url":null,"abstract":"This paper mainly uses the comparative analysis method and the case analysis method to explain the explainable artificial intelligence (XAI) in music. The use of XAI in music is real-time interactive, and the more explainable and transparent, the more accurate it is. Interpretability can occur in or after modeling, and deep learning provides theoretical support for XAI. Professor Nick Bryan-Kinns' team made a breakthrough through experimental research on “XAI+Music” with the difficulties currently being explained by artificial intelligence, XAI technology has been directly applied to the creative AI model, for “XAI+Music” development and innovation provide references and ideas.","PeriodicalId":398228,"journal":{"name":"2022 6th Asian Conference on Artificial Intelligence Technology (ACAIT)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113971524","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：