The segmentation of vehicles is a non-linear problem that has been tackled using methods for background subtraction in systems for traffic control. Probabilistic models, such as Gaussian Mixture Models (GMM), estimate the background of dynamic environments in this approach. The general modeling considers independent distributions for each pixel of the image. So, the classification is performed singly. The system uses often only one threshold to classify the pixels into background and foreground regions. This approach doest not work well when the cluster intersection is significant. In the vehicle segmentation, the color of the vehicles are similar to background, so the accuracy is affected. This paper proposes an approach to improve the classification of traffic scenes. This approach uses local thresholds to encourage the segmentation of vehicle regions. These thresholds are estimated by a spatial analysis of the previous classification. The results of the experiment performed shown that the classification process is improved by this approach.
{"title":"Adaptive Method for Segmentation of Vehicles through Local Threshold in the Gaussian Mixture Model","authors":"K. A. B. Lima, K. Aires, F. Reis","doi":"10.1109/BRACIS.2015.33","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.33","url":null,"abstract":"The segmentation of vehicles is a non-linear problem that has been tackled using methods for background subtraction in systems for traffic control. Probabilistic models, such as Gaussian Mixture Models (GMM), estimate the background of dynamic environments in this approach. The general modeling considers independent distributions for each pixel of the image. So, the classification is performed singly. The system uses often only one threshold to classify the pixels into background and foreground regions. This approach doest not work well when the cluster intersection is significant. In the vehicle segmentation, the color of the vehicles are similar to background, so the accuracy is affected. This paper proposes an approach to improve the classification of traffic scenes. This approach uses local thresholds to encourage the segmentation of vehicle regions. These thresholds are estimated by a spatial analysis of the previous classification. The results of the experiment performed shown that the classification process is improved by this approach.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128994181","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}
Alison R. Panisson, Felipe Meneguzzi, R. Vieira, Rafael Heitor Bordini
Argumentation is a key technique for reaching agreements in multi-agent systems. However, there are few practical approaches to develop multi-agent systems where agents engage in argumentation-based dialogues. In this paper, we give formal semantics to speech acts for argumentation-based dialogues in the context of an agent-oriented programming language. Our approach uses operational semantics and builds upon existing work that provides computationally grounded semantics for agent mental attitudes such as beliefs and goals. The paper also shows how our formal semantics can be used to prove properties of argumentation in multi-agent systems with direct reference to mental attitudes. We do so with an example of a proof sketch of termination of multi-agent dialogues under certain assumptions.
{"title":"Towards Practical Argumentation in Multi-agent Systems","authors":"Alison R. Panisson, Felipe Meneguzzi, R. Vieira, Rafael Heitor Bordini","doi":"10.1109/BRACIS.2015.30","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.30","url":null,"abstract":"Argumentation is a key technique for reaching agreements in multi-agent systems. However, there are few practical approaches to develop multi-agent systems where agents engage in argumentation-based dialogues. In this paper, we give formal semantics to speech acts for argumentation-based dialogues in the context of an agent-oriented programming language. Our approach uses operational semantics and builds upon existing work that provides computationally grounded semantics for agent mental attitudes such as beliefs and goals. The paper also shows how our formal semantics can be used to prove properties of argumentation in multi-agent systems with direct reference to mental attitudes. We do so with an example of a proof sketch of termination of multi-agent dialogues under certain assumptions.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127942842","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}
R. M. Santos, L. Matos, Hendrik T. Macedo, J. Filho
One of the biggest challenges in speech recognition today is its use on a daily basis, in which distortion and noise in the environment are present and hinder the recognition task. In the last thirty years, hundreds of methods for noise-robust recognition were proposed, each with its own advantages and disadvantages. In this paper, the use of convolutional neural networks (CNN) as acoustic models in automatic speech recognition systems (ASR) is proposed as an alternative to the classical recognition methods based on HMM without any noise-robust method applied. The experiment showed that the presented method reduces the equal error rate in word recognition tasks with additive noise.
{"title":"Speech Recognition in Noisy Environments with Convolutional Neural Networks","authors":"R. M. Santos, L. Matos, Hendrik T. Macedo, J. Filho","doi":"10.1109/BRACIS.2015.44","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.44","url":null,"abstract":"One of the biggest challenges in speech recognition today is its use on a daily basis, in which distortion and noise in the environment are present and hinder the recognition task. In the last thirty years, hundreds of methods for noise-robust recognition were proposed, each with its own advantages and disadvantages. In this paper, the use of convolutional neural networks (CNN) as acoustic models in automatic speech recognition systems (ASR) is proposed as an alternative to the classical recognition methods based on HMM without any noise-robust method applied. The experiment showed that the presented method reduces the equal error rate in word recognition tasks with additive noise.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125297990","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}
Constant optimization in symbolic regression is an important task addressed by several researchers. It has been demonstrated that continuous optimization techniques are adequate to find good values for the constants by minimizing the prediction error. In this paper, we evaluate several continuous optimization methods that can be used to perform constant optimization in symbolic regression. We have selected 14 well-known benchmark problems and tested the performance of diverse optimization methods in finding the expected constant values, assuming that the correct formula has been found. The results show that Levenberg-Marquardt presented the highest success rate among the evaluated methods, followed by Powell's and Nelder-Mead's Simplex. However, two benchmark problems were not solved, and for two other problems the Levenberg-Marquardt was largely outperformed by Nelder-Mead Simplex in terms of success rate. We conclude that even though a symbolic regression technique may find the correct formula, constant optimization may fail, thus, this may also happen during the search for a formula and may guide the method towards the wrong solution. Also, the efficiency of LM in finding high-quality solutions by using only a few function evaluations could serve as inspiration for the development of better symbolic regression methods.
{"title":"Evaluating Methods for Constant Optimization of Symbolic Regression Benchmark Problems","authors":"V. V. D. Melo, Benjamin Fowler, W. Banzhaf","doi":"10.1109/BRACIS.2015.55","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.55","url":null,"abstract":"Constant optimization in symbolic regression is an important task addressed by several researchers. It has been demonstrated that continuous optimization techniques are adequate to find good values for the constants by minimizing the prediction error. In this paper, we evaluate several continuous optimization methods that can be used to perform constant optimization in symbolic regression. We have selected 14 well-known benchmark problems and tested the performance of diverse optimization methods in finding the expected constant values, assuming that the correct formula has been found. The results show that Levenberg-Marquardt presented the highest success rate among the evaluated methods, followed by Powell's and Nelder-Mead's Simplex. However, two benchmark problems were not solved, and for two other problems the Levenberg-Marquardt was largely outperformed by Nelder-Mead Simplex in terms of success rate. We conclude that even though a symbolic regression technique may find the correct formula, constant optimization may fail, thus, this may also happen during the search for a formula and may guide the method towards the wrong solution. Also, the efficiency of LM in finding high-quality solutions by using only a few function evaluations could serve as inspiration for the development of better symbolic regression methods.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114357622","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}
This paper gives a batch SOM algorithm that is able to training a Kohonen map taking into account simultaneously several dissimilarity matrices, that are obtained using different sets of variables and dissimilarity functions. This algorithm is designed to provide a partition and a set-medoids vector representative for each cluster, and learn a relevance weight on the training for each dissimilarity matrix by optimizing an objective function. These relevance weights change at each algorithm's iteration and are different from one cluster to another. The proposed algorithm provides a collaborative role of the different dissimilarity matrices, aiming to cluster and visualizing the data while preserving their topology. Several examples illustrate the usefulness of the proposed algorithm.
{"title":"A Set-Medoids Vector Batch SOM Algorithm Based on Multiple Dissimilarity Matrices","authors":"F. D. Carvalho, Eduardo C. Simões","doi":"10.1109/BRACIS.2015.13","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.13","url":null,"abstract":"This paper gives a batch SOM algorithm that is able to training a Kohonen map taking into account simultaneously several dissimilarity matrices, that are obtained using different sets of variables and dissimilarity functions. This algorithm is designed to provide a partition and a set-medoids vector representative for each cluster, and learn a relevance weight on the training for each dissimilarity matrix by optimizing an objective function. These relevance weights change at each algorithm's iteration and are different from one cluster to another. The proposed algorithm provides a collaborative role of the different dissimilarity matrices, aiming to cluster and visualizing the data while preserving their topology. Several examples illustrate the usefulness of the proposed algorithm.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114569330","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}
Noise filtering can be considered an important pre-processing step in the data mining process, making data more reliable for pattern extraction. An interesting aspect for increasing data understanding would be to rank the potential noisy cases, in order to evidence the most unreliable instances to be further examined. Since the majority of the filters from the literature were designed only for hard classification, distinguishing whether an example is noisy or not, in this paper we adapt the output of some state of the art noise filters for ranking the cases identified as suspicious. We also present new evaluation measures for the noise rankers designed, which take into account the ordering of the detected noisy cases.
{"title":"Adapting Noise Filters for Ranking","authors":"Ana Carolina Lorena, L. P. F. Garcia, A. Carvalho","doi":"10.1109/BRACIS.2015.58","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.58","url":null,"abstract":"Noise filtering can be considered an important pre-processing step in the data mining process, making data more reliable for pattern extraction. An interesting aspect for increasing data understanding would be to rank the potential noisy cases, in order to evidence the most unreliable instances to be further examined. Since the majority of the filters from the literature were designed only for hard classification, distinguishing whether an example is noisy or not, in this paper we adapt the output of some state of the art noise filters for ranking the cases identified as suspicious. We also present new evaluation measures for the noise rankers designed, which take into account the ordering of the detected noisy cases.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134463494","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}
Sentiment Analysis is the field of study that analyzes people's opinions in texts. In the last decade, humans have come to share their opinions in social media on the Web (e.g., Forum discussions and posts in social network sites). Opinions are important because whenever we need to take a decision, we want to know others' points of view. The interest of industry and academia in this field of study is partly due to its potential applications, such as: marketing, public relations and political campaign. Research in this field often considers English data, while data from other languages are less explored. In this work we evaluate the available resources to assist Portuguese language sentiment analysis. For doing this, we perform sentiment analysis in a data set of the accommodation sector. We compare different pos-taggers and sentiment lexicons. We also evaluate the impact of some linguistic rules regarding negation and the position of adjectives.
{"title":"Exploring Resources for Sentiment Analysis in Portuguese Language","authors":"L. Freitas, R. Vieira","doi":"10.1109/BRACIS.2015.52","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.52","url":null,"abstract":"Sentiment Analysis is the field of study that analyzes people's opinions in texts. In the last decade, humans have come to share their opinions in social media on the Web (e.g., Forum discussions and posts in social network sites). Opinions are important because whenever we need to take a decision, we want to know others' points of view. The interest of industry and academia in this field of study is partly due to its potential applications, such as: marketing, public relations and political campaign. Research in this field often considers English data, while data from other languages are less explored. In this work we evaluate the available resources to assist Portuguese language sentiment analysis. For doing this, we perform sentiment analysis in a data set of the accommodation sector. We compare different pos-taggers and sentiment lexicons. We also evaluate the impact of some linguistic rules regarding negation and the position of adjectives.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128076544","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}
Stealth-based path finding focuses on finding a path between two points that minimizes the agent's exposure to patrolling units. This problem arises in the robotics field and in modern games, in which an agent must traverse the terrain covertly. This paper presents a real-time method capable of finding a covert path in a terrain patrolled by multiple moving agents using a special navigation mesh. The generated path passes through cover whenever possible in order to avoid open areas and reduce its overall visibility. Also, the stealthy agent uses different movement speed based on the current area it traverses. It is considered here that the environment is known and static.
{"title":"Stealthy Path Planning Using Navigation Meshes","authors":"Matheus R. F. Mendonça, H. Bernardino, R. F. Neto","doi":"10.1109/BRACIS.2015.49","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.49","url":null,"abstract":"Stealth-based path finding focuses on finding a path between two points that minimizes the agent's exposure to patrolling units. This problem arises in the robotics field and in modern games, in which an agent must traverse the terrain covertly. This paper presents a real-time method capable of finding a covert path in a terrain patrolled by multiple moving agents using a special navigation mesh. The generated path passes through cover whenever possible in order to avoid open areas and reduce its overall visibility. Also, the stealthy agent uses different movement speed based on the current area it traverses. It is considered here that the environment is known and static.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122106489","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}
Ricardo M. S. Budaruiche, Renata Wassermann, D. Patrão, M. I. Achatz
The computer-assisted search for knowledge in the medical field has become increasingly frequent. Scientific progress in subjects such as ontology and artificial intelligence allowed researchers to develop methods for capturing, using and sharing specific knowledge. The Li-Fraumeni Syndrome (LFS) is a syndrome that causes multiple primary tumors in children and young adults. These tumors can be Breast Cancer, Brain Tumors and Sarcomas, among others. This paper presents a case study of a new set of ontologies in the domain of the LFS with the objective of extracting knowledge about patients that fit clinical criteria of one or more of the four LFS clinical criteria: Classic, Birch, Eeles and Chompret.
{"title":"Li-Fraumeni Ontology: A Case Study of an Ontology for Knowledge Discovery in a Cancer Domain","authors":"Ricardo M. S. Budaruiche, Renata Wassermann, D. Patrão, M. I. Achatz","doi":"10.1109/BRACIS.2015.41","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.41","url":null,"abstract":"The computer-assisted search for knowledge in the medical field has become increasingly frequent. Scientific progress in subjects such as ontology and artificial intelligence allowed researchers to develop methods for capturing, using and sharing specific knowledge. The Li-Fraumeni Syndrome (LFS) is a syndrome that causes multiple primary tumors in children and young adults. These tumors can be Breast Cancer, Brain Tumors and Sarcomas, among others. This paper presents a case study of a new set of ontologies in the domain of the LFS with the objective of extracting knowledge about patients that fit clinical criteria of one or more of the four LFS clinical criteria: Classic, Birch, Eeles and Chompret.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121943098","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}
Eden P. da Silva, C. Estombelo-Montesco, E. Santana
Machine learning algorithms are used in many areas, in signal processing, the adaptive filtering has been used in many jobs as smooth, prediction, equalization, etc. The Least Mean Square (LMS) algorithm is a successful example of this approach, this algorithm takes the instantaneous gradient of the cost function in his learning process. Nevertheless, recent works have proposed improvements on adaptive filtering based in LMS. On this context, Sigmoid Algorithm changes the LMS cost function, the Mean Square Error, to an even error function, which improves the convergence rate on the learning process. On a more complex approach, the kernel LMS taking the filtering problem in a high dimensional Hilbert space generated for a kernel function, where the desired filter output is the result of algebraic operations in that kernel generated space, which resulted on a decrease of the error compared to LMS. In face of this two improvements, this paper describes our work propose, the kernel version of Sigmoid Algorithm, whose results showed a decrease in the convergence rate on the learning process compared to kernel LMS.
{"title":"KSIG: Improving the Convergence Rate in Adaptive Filtering Using Kernel Hilbert Space","authors":"Eden P. da Silva, C. Estombelo-Montesco, E. Santana","doi":"10.1109/BRACIS.2015.54","DOIUrl":"https://doi.org/10.1109/BRACIS.2015.54","url":null,"abstract":"Machine learning algorithms are used in many areas, in signal processing, the adaptive filtering has been used in many jobs as smooth, prediction, equalization, etc. The Least Mean Square (LMS) algorithm is a successful example of this approach, this algorithm takes the instantaneous gradient of the cost function in his learning process. Nevertheless, recent works have proposed improvements on adaptive filtering based in LMS. On this context, Sigmoid Algorithm changes the LMS cost function, the Mean Square Error, to an even error function, which improves the convergence rate on the learning process. On a more complex approach, the kernel LMS taking the filtering problem in a high dimensional Hilbert space generated for a kernel function, where the desired filter output is the result of algebraic operations in that kernel generated space, which resulted on a decrease of the error compared to LMS. In face of this two improvements, this paper describes our work propose, the kernel version of Sigmoid Algorithm, whose results showed a decrease in the convergence rate on the learning process compared to kernel LMS.","PeriodicalId":416771,"journal":{"name":"2015 Brazilian Conference on Intelligent Systems (BRACIS)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122217690","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：