Pub Date : 2020-09-01DOI: 10.1109/TransAI49837.2020.00008
Tsung-Min Huang, Hunter Hsieh, Jiaqi Qin, Hsien-Fung Liu, M. Eirinaki
Relating sounds to visuals, like photographs, is something humans do subconsciously every day. Deep learning has allowed for several image-related applications, with some focusing on generating labels for images, or synthesize images from a text description. Similarly, it has been employed to create new music scores from existing ones, or add lyrics to a song. In this work, we bring sight and sound together and present IMuCo, an intelligent music composer that creates original music for any given image, taking into consideration what its implied mood is. Our music augmentation and composing methodology attempts to translate image “linguistics” into music “linguistics” without any intermediate natural language translation steps. We propose an encoder-decoder architecture to translate an image into music, first classifying it into one of predefined moods, then generating music to match it. We discuss in detail how we created the training dataset, including several feature engineering decisions in terms of representing music. We also introduce an evaluation classifier framework used for validation and evaluation of the system, and present experimental results of IMuCo’s prototype for two moods: happy and sad. IMuCo can be the core component of a framework that composes the soundtrack for longer video clips, used in advertising, art, and entertainment industries.
{"title":"Play it again IMuCo! Music Composition to Match your Mood","authors":"Tsung-Min Huang, Hunter Hsieh, Jiaqi Qin, Hsien-Fung Liu, M. Eirinaki","doi":"10.1109/TransAI49837.2020.00008","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00008","url":null,"abstract":"Relating sounds to visuals, like photographs, is something humans do subconsciously every day. Deep learning has allowed for several image-related applications, with some focusing on generating labels for images, or synthesize images from a text description. Similarly, it has been employed to create new music scores from existing ones, or add lyrics to a song. In this work, we bring sight and sound together and present IMuCo, an intelligent music composer that creates original music for any given image, taking into consideration what its implied mood is. Our music augmentation and composing methodology attempts to translate image “linguistics” into music “linguistics” without any intermediate natural language translation steps. We propose an encoder-decoder architecture to translate an image into music, first classifying it into one of predefined moods, then generating music to match it. We discuss in detail how we created the training dataset, including several feature engineering decisions in terms of representing music. We also introduce an evaluation classifier framework used for validation and evaluation of the system, and present experimental results of IMuCo’s prototype for two moods: happy and sad. IMuCo can be the core component of a framework that composes the soundtrack for longer video clips, used in advertising, art, and entertainment industries.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"159 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123382365","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00017
Joseph R. Barr, Peter Shaw, F. Abu-Khzam, Sheng Yu, Heng Yin, Tyler Thatcher
Empirical analysis of source code of Android Fluoride Bluetooth stack demonstrates a novel approach of classification of source code and rating for vulnerability. A workflow that combines deep learning and combinatorial techniques with a straightforward random forest regression is presented. Two kinds of embedding are used: code2vec and LSTM, resulting in a distance matrix that is interpreted as a (combinatorial) graph whose vertices represent code components, functions and methods. Cluster Editing is then applied to partition the vertex set of the graph into subsets representing nearly complete subgraphs. Finally, the vectors representing the components are used as features to model the components for vulnerability risk.
{"title":"Combinatorial Code Classification & Vulnerability Rating","authors":"Joseph R. Barr, Peter Shaw, F. Abu-Khzam, Sheng Yu, Heng Yin, Tyler Thatcher","doi":"10.1109/TransAI49837.2020.00017","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00017","url":null,"abstract":"Empirical analysis of source code of Android Fluoride Bluetooth stack demonstrates a novel approach of classification of source code and rating for vulnerability. A workflow that combines deep learning and combinatorial techniques with a straightforward random forest regression is presented. Two kinds of embedding are used: code2vec and LSTM, resulting in a distance matrix that is interpreted as a (combinatorial) graph whose vertices represent code components, functions and methods. Cluster Editing is then applied to partition the vertex set of the graph into subsets representing nearly complete subgraphs. Finally, the vectors representing the components are used as features to model the components for vulnerability risk.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121173260","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00024
Golnoush Asaeikheybari, Cory Hughart, Devansh Gupta, A. Avery, Mary M. Step, Jennifer McMillen Smith, Joshua Kratz, Julia Briggs, Ming-chun Huang
Mobile phone applications provide a new and easy-access platform for delivering tailored human immunodeficiency virus (HIV) and sexually transmitted disease (STD) prevention and care. Recent researches have shown that mobile interventions have positive effects in adhesive to care program, antiretroviral therapy (ART), self-management of disease, and are also critical in decreasing the HIV pandemic, and stigmatization. In this paper, a precision health app, Positive Peers (PP), has been developed collaboratively while enabled by data harnessing, Artificial Intelligence (Al), and learning. Positive Peers is an Android/iOS-based social media app for providing support and information to a young adult subgroup living with HIV who are in strong need of support and motivation. We apply an intervention approach combined with Natural Language Processing (NLP) to help the targeted youth to engage more with the app. Using NLP facilitates the flow of information that has a critical role in decreasing the uncertainty of patients by being injected to useful related information. It further improves the interaction of users of the app while providing a compact platform for users to better find the answers to their questions and concerns. The NLP system has been evaluated in an alpha test.
{"title":"Precision HIV Health App, Positive Peers, Powered by Data Harnessing, AI, and Learning","authors":"Golnoush Asaeikheybari, Cory Hughart, Devansh Gupta, A. Avery, Mary M. Step, Jennifer McMillen Smith, Joshua Kratz, Julia Briggs, Ming-chun Huang","doi":"10.1109/TransAI49837.2020.00024","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00024","url":null,"abstract":"Mobile phone applications provide a new and easy-access platform for delivering tailored human immunodeficiency virus (HIV) and sexually transmitted disease (STD) prevention and care. Recent researches have shown that mobile interventions have positive effects in adhesive to care program, antiretroviral therapy (ART), self-management of disease, and are also critical in decreasing the HIV pandemic, and stigmatization. In this paper, a precision health app, Positive Peers (PP), has been developed collaboratively while enabled by data harnessing, Artificial Intelligence (Al), and learning. Positive Peers is an Android/iOS-based social media app for providing support and information to a young adult subgroup living with HIV who are in strong need of support and motivation. We apply an intervention approach combined with Natural Language Processing (NLP) to help the targeted youth to engage more with the app. Using NLP facilitates the flow of information that has a critical role in decreasing the uncertainty of patients by being injected to useful related information. It further improves the interaction of users of the app while providing a compact platform for users to better find the answers to their questions and concerns. The NLP system has been evaluated in an alpha test.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"166 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116304303","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00021
M. A. El-Wahab, F. Abu-Khzam, Kai Wang, Peter Shaw
The last decade witnessed an increased interest in exact and parameterized exponential-time algorithms for NP - hard problems. The hardness of polynomial-time approximation of many intractable problems motivated the work on fixed-parameter approximation where polynomial-time is relaxed into FPT -time as long as improved approximation is obtained, most often requiring constant ratio bounds. In this paper we move a step further by investigating the practicality of exponential time approximation (versus FPT-time) as long as obtained solutions are within an additive parameter. The running time of such algorithm would be reduced by some function (factor) of the same parameter. The objective is to obtain a cost-effective trade-off between reduced running time and quality of approximation while providing provably near optimal solutions. We present experimental studies of two problems: Dominating Set and Vertex Cover. Our experiments show that semi-exact algorithms are indeed very promising.
{"title":"Semi-Exact Exponential-Time Algorithms: an Experimental Study","authors":"M. A. El-Wahab, F. Abu-Khzam, Kai Wang, Peter Shaw","doi":"10.1109/TransAI49837.2020.00021","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00021","url":null,"abstract":"The last decade witnessed an increased interest in exact and parameterized exponential-time algorithms for NP - hard problems. The hardness of polynomial-time approximation of many intractable problems motivated the work on fixed-parameter approximation where polynomial-time is relaxed into FPT -time as long as improved approximation is obtained, most often requiring constant ratio bounds. In this paper we move a step further by investigating the practicality of exponential time approximation (versus FPT-time) as long as obtained solutions are within an additive parameter. The running time of such algorithm would be reduced by some function (factor) of the same parameter. The objective is to obtain a cost-effective trade-off between reduced running time and quality of approximation while providing provably near optimal solutions. We present experimental studies of two problems: Dominating Set and Vertex Cover. Our experiments show that semi-exact algorithms are indeed very promising.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"111 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131951193","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00022
Joseph R. Barr, Peter Shaw, F. Abu-Khzam
Every graph has orientation $delta$ with the property that the indegree and outdegree of each vertex differ by no more than a unity. For a subset A of vertices of a digraph D the indegree of A is the number of arcs pointing into A and the outdegree of A is the number of arcs pointing out of A. The flux at A is the difference of the two (‘in’ minus ‘out’.) For a fixed graph G consider the set $triangle$ of all orientations of G. We calculate “worstcase” flux as the “min-max” flux: the maximum flux over all subsets of vertices and the minimum over all orientations. The min-max flux over A with respect to orientation $delta$ is the “flux” of the graph $phi_{delta}(A)$ wherebegin{equation*}min_{deltaindelta A}max_{subset V}phi(A;delta). tag{1}end{equation*}An orientation $delta$ of G achieving the min-max is said to be optimally-balanced. In this paper we characterize optimally-balanced graphs.
{"title":"Optimally Balanced Orientation of Graphs","authors":"Joseph R. Barr, Peter Shaw, F. Abu-Khzam","doi":"10.1109/TransAI49837.2020.00022","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00022","url":null,"abstract":"Every graph has orientation $delta$ with the property that the indegree and outdegree of each vertex differ by no more than a unity. For a subset A of vertices of a digraph D the indegree of A is the number of arcs pointing into A and the outdegree of A is the number of arcs pointing out of A. The flux at A is the difference of the two (‘in’ minus ‘out’.) For a fixed graph G consider the set $triangle$ of all orientations of G. We calculate “worstcase” flux as the “min-max” flux: the maximum flux over all subsets of vertices and the minimum over all orientations. The min-max flux over A with respect to orientation $delta$ is the “flux” of the graph $phi_{delta}(A)$ wherebegin{equation*}min_{deltaindelta A}max_{subset V}phi(A;delta). tag{1}end{equation*}An orientation $delta$ of G achieving the min-max is said to be optimally-balanced. In this paper we characterize optimally-balanced graphs.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132967982","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}
Recently, human pose estimation has received much attention in the research community due to its broad range of application scenarios. Most architectures for human pose estimation use multiple resolution networks, such as Hourglass, CPN, HRNet, etc. High Resolution Network (HRNet) is the latest SOTA architecture improved from Hourglass. In this paper, we propose a novel attention block that leverages a special Channel-Attention branch. We use this attention block as the building block and adopt the architecture of HRNet to build our Attention Based HRNet (A-HRNet). Experiments show that our model can consistently outperform HRNet on different datasets. Moreover, our model achieves the state-of-the-art performance on the COCO keypoint detection val2017 dataset (77.7 AP)1.
{"title":"A-HRNet: Attention Based High Resolution Network for Human pose estimation","authors":"Ying Li, Chenxi Wang, Yu Cao, Benyuan Liu, Yan Luo, Honggang Zhang","doi":"10.1109/TransAI49837.2020.00016","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00016","url":null,"abstract":"Recently, human pose estimation has received much attention in the research community due to its broad range of application scenarios. Most architectures for human pose estimation use multiple resolution networks, such as Hourglass, CPN, HRNet, etc. High Resolution Network (HRNet) is the latest SOTA architecture improved from Hourglass. In this paper, we propose a novel attention block that leverages a special Channel-Attention branch. We use this attention block as the building block and adopt the architecture of HRNet to build our Attention Based HRNet (A-HRNet). Experiments show that our model can consistently outperform HRNet on different datasets. Moreover, our model achieves the state-of-the-art performance on the COCO keypoint detection val2017 dataset (77.7 AP)1.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114935002","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 : 2020-09-01DOI: 10.1109/transai49837.2020.00003
{"title":"[Copyright nnotice]","authors":"","doi":"10.1109/transai49837.2020.00003","DOIUrl":"https://doi.org/10.1109/transai49837.2020.00003","url":null,"abstract":"","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121469312","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00023
Julian Vu, Katerina Potika
Computing the edge betweenness centrality is an important step in a great deal of the analysis tasks of community structures in complex networks. It mostly serves as a measure for the traffic or flow of a particular edge in connecting various parts or communities together. Various algorithms that compute the edge betweenness centrality in general graphs exist but they are expensive. In this paper, we design an algorithm that takes advantage of the structure of tree graphs to compute the edge betweenness centrality more efficiently in such graphs and perform experiments on random graphs.
{"title":"Edge Betweenness Centrality on Trees","authors":"Julian Vu, Katerina Potika","doi":"10.1109/TransAI49837.2020.00023","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00023","url":null,"abstract":"Computing the edge betweenness centrality is an important step in a great deal of the analysis tasks of community structures in complex networks. It mostly serves as a measure for the traffic or flow of a particular edge in connecting various parts or communities together. Various algorithms that compute the edge betweenness centrality in general graphs exist but they are expensive. In this paper, we design an algorithm that takes advantage of the structure of tree graphs to compute the edge betweenness centrality more efficiently in such graphs and perform experiments on random graphs.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134244900","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00028
Anousha Athreya, S. K. Taswell, Sohyb Mashkoor, C. Taswell
We discuss definitions of entities, equality, and equivalence as used by a transdisciplinary diversity of research fields including mathematics, statistics, computational linguistics, computer programming, knowledge engineering, and music theory. Declaring definitions for these concepts in the situational context of each domain specific field supports the essential question ‘Equal or equivalent entities?’ about two things as same, similar, related, or different for that field. Pattern recognition performed by artificial intelligence applications can be described as the automated process of answering this fundamental question about the similarity or difference between two things.
{"title":"Essential Question: ‘Equal or Equivalent Entities?’ About Two Things as Same, Similar, or Different","authors":"Anousha Athreya, S. K. Taswell, Sohyb Mashkoor, C. Taswell","doi":"10.1109/TransAI49837.2020.00028","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00028","url":null,"abstract":"We discuss definitions of entities, equality, and equivalence as used by a transdisciplinary diversity of research fields including mathematics, statistics, computational linguistics, computer programming, knowledge engineering, and music theory. Declaring definitions for these concepts in the situational context of each domain specific field supports the essential question ‘Equal or equivalent entities?’ about two things as same, similar, related, or different for that field. Pattern recognition performed by artificial intelligence applications can be described as the automated process of answering this fundamental question about the similarity or difference between two things.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"54 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130781817","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 : 2020-09-01DOI: 10.1109/TransAI49837.2020.00033
Maria Krommyda, Verena Kantere
As the Internet of Things (IoT) systems gain in popularity, an increasing number of Big Data sources are available. Ranging from small sensor networks designed for household use to large fully automated industrial environments, the Internet of Things systems create billions of measurements each second making traditional storage and indexing solutions obsolete. While research around Big Data has focused on scalable solutions that can support the datasets produced by these systems, the focus has been mainly on managing the volume and velocity of these data, rather than providing efficient solutions for their retrieval and analysis. A key characteristic of these data, which is, more often than not, overlooked, is the spatial information that can be used to integrate data from multiple sources and conduct multidimensional analysis of the collected information. We present here the solutions currently available for the storage and indexing of spatial datasets produced by the IoT systems and we discuss their applicability in real world scenarios.
{"title":"Spatial Data Management in IoT systems: A study of available storage and indexing solutions","authors":"Maria Krommyda, Verena Kantere","doi":"10.1109/TransAI49837.2020.00033","DOIUrl":"https://doi.org/10.1109/TransAI49837.2020.00033","url":null,"abstract":"As the Internet of Things (IoT) systems gain in popularity, an increasing number of Big Data sources are available. Ranging from small sensor networks designed for household use to large fully automated industrial environments, the Internet of Things systems create billions of measurements each second making traditional storage and indexing solutions obsolete. While research around Big Data has focused on scalable solutions that can support the datasets produced by these systems, the focus has been mainly on managing the volume and velocity of these data, rather than providing efficient solutions for their retrieval and analysis. A key characteristic of these data, which is, more often than not, overlooked, is the spatial information that can be used to integrate data from multiple sources and conduct multidimensional analysis of the collected information. We present here the solutions currently available for the storage and indexing of spatial datasets produced by the IoT systems and we discuss their applicability in real world scenarios.","PeriodicalId":151527,"journal":{"name":"2020 Second International Conference on Transdisciplinary AI (TransAI)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116021700","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：