Borja Bordel, R. Alcarria, Joaquín Chung, R. Kettimuthu
Future Industry 4.0 scenarios are characterized by seamless integration between computational and physical processes. To achieve this objective, dense platforms made of small sensing nodes and other resource constraint devices are ubiquitously deployed. All these devices have a limited number of computational resources, just enough to perform the simple operation they are in charge of. The remaining operations are delegated to powerful gateways that manage sensing nodes, but resources are never unlimited, and as more and more devices are deployed on Industry 4.0 platforms, gateways present more problems to handle massive machine-type communications. Although the problems are diverse, those related to security are especially critical. To enable sensing nodes to establish secure communications, several semiconductor companies are currently promoting a new generation of devices based on Physical Unclonable Functions, whose usage grows every year in many real industrial scenarios. Those hardware devices do not consume any computational resource but force the gateway to keep large key-value catalogues for each individual node. In this context, memory usage is not scalable and processing delays increase exponentially with each new node on the platform. In this paper, we address this challenge through predictor-corrector models, representing the key-value catalogues. Models are mathematically complex, but we argue that they consume less computational resources than current approaches. The lightweight models are based on complex functions managed as Laurent series, cubic spline interpolations, and Boolean functions also developed as series. Unknown parameters in these models are predicted, and eventually corrected to calculate the output value for each given key. The initial parameters are based on the Kane Yee formula. An experimental analysis and a performance evaluation are provided in the experimental section, showing that the proposed approach causes a significant reduction in the resource consumption.
{"title":"Predictor-corrector models for lightweight massive machine-type communications in Industry 4.0","authors":"Borja Bordel, R. Alcarria, Joaquín Chung, R. Kettimuthu","doi":"10.3233/ica-230713","DOIUrl":"https://doi.org/10.3233/ica-230713","url":null,"abstract":"Future Industry 4.0 scenarios are characterized by seamless integration between computational and physical processes. To achieve this objective, dense platforms made of small sensing nodes and other resource constraint devices are ubiquitously deployed. All these devices have a limited number of computational resources, just enough to perform the simple operation they are in charge of. The remaining operations are delegated to powerful gateways that manage sensing nodes, but resources are never unlimited, and as more and more devices are deployed on Industry 4.0 platforms, gateways present more problems to handle massive machine-type communications. Although the problems are diverse, those related to security are especially critical. To enable sensing nodes to establish secure communications, several semiconductor companies are currently promoting a new generation of devices based on Physical Unclonable Functions, whose usage grows every year in many real industrial scenarios. Those hardware devices do not consume any computational resource but force the gateway to keep large key-value catalogues for each individual node. In this context, memory usage is not scalable and processing delays increase exponentially with each new node on the platform. In this paper, we address this challenge through predictor-corrector models, representing the key-value catalogues. Models are mathematically complex, but we argue that they consume less computational resources than current approaches. The lightweight models are based on complex functions managed as Laurent series, cubic spline interpolations, and Boolean functions also developed as series. Unknown parameters in these models are predicted, and eventually corrected to calculate the output value for each given key. The initial parameters are based on the Kane Yee formula. An experimental analysis and a performance evaluation are provided in the experimental section, showing that the proposed approach causes a significant reduction in the resource consumption.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69926714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
F. Segovia, J. Ramírez, D. Salas-González, I. A. Illán, Francisco J. Martínez-Murcia, J. Rodriguez-Rivero, F. J. Leiva, C. Gaitan, J. Górriz
The stable supply of electricity is essential for the industrial activity and economic development as well as for human welfare. For this reason, electrical system devices are equipped with monitoring systems that facilitate their management and ensure an uninterrupted operation. This is the case of electrical power transformers, which usually have monitoring systems that allow early detection of anomalies in order to prevent potential malfunctions. These monitoring systems typically make use of sensors that are in physical contact with the transformer devices and can therefore be affected by transformer problems. In this work we demonstrate a monitoring system for electrical power transformers based on temperature measurements obtained by means of thermal cameras. Properly positioned, the cameras provide thermal data of the transformer, the incoming and outgoing lines and their surroundings. Subsequently, by appropriate image processing, it is possible to obtain temperature series to monitor the transformer operation. In addition, the system stores and processes thermal data in external equipment (placed in locations other than the transformers) and is equipped with a communications module that allows secure data transmission independent of the power grid. This aspect, along with the fact that there is no need to have physical contact with the transformer, make this approach safer and more reliable than standard approaches based on sensors. The proposed system has been evaluated in 14 stations belonging to the Spanish power grid, obtaining accurate and reliable temperature time series.
{"title":"Connected system for monitoring electrical power transformers using thermal imaging","authors":"F. Segovia, J. Ramírez, D. Salas-González, I. A. Illán, Francisco J. Martínez-Murcia, J. Rodriguez-Rivero, F. J. Leiva, C. Gaitan, J. Górriz","doi":"10.3233/ica-230712","DOIUrl":"https://doi.org/10.3233/ica-230712","url":null,"abstract":"The stable supply of electricity is essential for the industrial activity and economic development as well as for human welfare. For this reason, electrical system devices are equipped with monitoring systems that facilitate their management and ensure an uninterrupted operation. This is the case of electrical power transformers, which usually have monitoring systems that allow early detection of anomalies in order to prevent potential malfunctions. These monitoring systems typically make use of sensors that are in physical contact with the transformer devices and can therefore be affected by transformer problems. In this work we demonstrate a monitoring system for electrical power transformers based on temperature measurements obtained by means of thermal cameras. Properly positioned, the cameras provide thermal data of the transformer, the incoming and outgoing lines and their surroundings. Subsequently, by appropriate image processing, it is possible to obtain temperature series to monitor the transformer operation. In addition, the system stores and processes thermal data in external equipment (placed in locations other than the transformers) and is equipped with a communications module that allows secure data transmission independent of the power grid. This aspect, along with the fact that there is no need to have physical contact with the transformer, make this approach safer and more reliable than standard approaches based on sensors. The proposed system has been evaluated in 14 stations belonging to the Spanish power grid, obtaining accurate and reliable temperature time series.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69926640","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
3D reconstruction is extremely important in CAD (computer-aided design)/CAE (computer-aided Engineering)/CAM (computer-aided manufacturing). For interpretability, reinforcement learning (RL) is used to reconstruct 3D shapes from images by a series of editing actions. However, typical applications of RL for 3D reconstruction face problems. The search space will increase exponentially with the action space due to the curse of dimensionality, which leads to low performance, especially for complex action spaces in 3D reconstruction. Additionally, most works involve training a specific agent for each shape class without learning related experiences from others. Therefore, we present a hierarchical RL approach with transferability to reconstruct 3D shapes (HRLT3D). First, actions are grouped into macro actions that can be chosen by the top-agent. Second, the task is accordingly decomposed into hierarchically simplified sub-tasks solved by sub-agents. Different from classical hierarchical RL (HRL), we propose a sub-agent based on augmented state space (ASS-Sub-Agent) to replace a set of sub-agents, which can speed up the training process due to shared learning and having fewer parameters. Furthermore, the ASS-Sub-Agent is more easily transferred to data of other classes due to the augmented diverse states and the simplified tasks. The experimental results on typical public dataset show that the proposed HRLT3D performs overwhelmingly better than recent baselines. More impressingly, the experiments also demonstrate the extreme transferability of our approach among data of different classes.
{"title":"3D reconstruction based on hierarchical reinforcement learning with transferability","authors":"Lan Li, Fazhi He, Rubin Fan, Bo Fan, Xiaohu Yan","doi":"10.3233/ica-230710","DOIUrl":"https://doi.org/10.3233/ica-230710","url":null,"abstract":"3D reconstruction is extremely important in CAD (computer-aided design)/CAE (computer-aided Engineering)/CAM (computer-aided manufacturing). For interpretability, reinforcement learning (RL) is used to reconstruct 3D shapes from images by a series of editing actions. However, typical applications of RL for 3D reconstruction face problems. The search space will increase exponentially with the action space due to the curse of dimensionality, which leads to low performance, especially for complex action spaces in 3D reconstruction. Additionally, most works involve training a specific agent for each shape class without learning related experiences from others. Therefore, we present a hierarchical RL approach with transferability to reconstruct 3D shapes (HRLT3D). First, actions are grouped into macro actions that can be chosen by the top-agent. Second, the task is accordingly decomposed into hierarchically simplified sub-tasks solved by sub-agents. Different from classical hierarchical RL (HRL), we propose a sub-agent based on augmented state space (ASS-Sub-Agent) to replace a set of sub-agents, which can speed up the training process due to shared learning and having fewer parameters. Furthermore, the ASS-Sub-Agent is more easily transferred to data of other classes due to the augmented diverse states and the simplified tasks. The experimental results on typical public dataset show that the proposed HRLT3D performs overwhelmingly better than recent baselines. More impressingly, the experiments also demonstrate the extreme transferability of our approach among data of different classes.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69926530","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Leandro Ruiz, S. Díaz, Jose M. Gonzalez, Francisco Cavas-Martínez
The accuracy and reliability requirements in aerospace manufacturing processes are some of the most demanding in industry. One of the first steps is detection and precise measurement using artificial vision models to accurately process the part. However, these systems require complex adjustments and do not work correctly in uncontrolled scenarios, but require manual supervision, which reduces the autonomy of automated machinery. To solve these problems, this paper proposes a convolutional neural network for the detection and measurement of drills and other fixation elements in an uncontrolled industrial manufacturing environment. In addition, a fine-tuning algorithm is applied to the results obtained from the network, and a new metric is defined to evaluate the quality of detection. The efficiency and robustness of the proposed method were verified in a real production environment, with 99.7% precision, 97.6% recall and an overall quality factor of 96.0%. The reduction in operator intervention went from 13.3% to 0.6%. The presented work will allow the competitiveness of aircraft component manufacturing processes to increase, and working environments will be safer and more efficient.
{"title":"Improving the competitiveness of aircraft manufacturing automated processes by a deep neural network","authors":"Leandro Ruiz, S. Díaz, Jose M. Gonzalez, Francisco Cavas-Martínez","doi":"10.3233/ica-230711","DOIUrl":"https://doi.org/10.3233/ica-230711","url":null,"abstract":"The accuracy and reliability requirements in aerospace manufacturing processes are some of the most demanding in industry. One of the first steps is detection and precise measurement using artificial vision models to accurately process the part. However, these systems require complex adjustments and do not work correctly in uncontrolled scenarios, but require manual supervision, which reduces the autonomy of automated machinery. To solve these problems, this paper proposes a convolutional neural network for the detection and measurement of drills and other fixation elements in an uncontrolled industrial manufacturing environment. In addition, a fine-tuning algorithm is applied to the results obtained from the network, and a new metric is defined to evaluate the quality of detection. The efficiency and robustness of the proposed method were verified in a real production environment, with 99.7% precision, 97.6% recall and an overall quality factor of 96.0%. The reduction in operator intervention went from 13.3% to 0.6%. The presented work will allow the competitiveness of aircraft component manufacturing processes to increase, and working environments will be safer and more efficient.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69926568","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Rosso, A. Aloisio, V. Randazzo, L. Tanzi, G. Cirrincione, G. Marano
In the last decades, the majority of the existing infrastructure heritage is approaching the end of its nominal design life mainly due to aging, deterioration, and degradation phenomena, threatening the safety levels of these strategic routes of communications. For civil engineers and researchers devoted to assessing and monitoring the structural health (SHM) of existing structures, the demand for innovative indirect non-destructive testing (NDT) methods aided with artificial intelligence (AI) is progressively spreading. In the present study, the authors analyzed the exertion of various deep learning models in order to increase the productivity of classifying ground penetrating radar (GPR) images for SHM purposes, especially focusing on road tunnel linings evaluations. Specifically, the authors presented a comparative study employing two convolutional models, i.e. the ResNet-50 and the EfficientNet-B0, and a recent transformer model, i.e. the Vision Transformer (ViT). Precisely, the authors evaluated the effects of training the models with or without pre-processed data through the bi-dimensional Fourier transform. Despite the theoretical advantages envisaged by adopting this kind of pre-processing technique on GPR images, the best classification performances have been still manifested by the classifiers trained without the Fourier pre-processing.
{"title":"Comparative deep learning studies for indirect tunnel monitoring with and without Fourier pre-processing","authors":"M. Rosso, A. Aloisio, V. Randazzo, L. Tanzi, G. Cirrincione, G. Marano","doi":"10.3233/ica-230709","DOIUrl":"https://doi.org/10.3233/ica-230709","url":null,"abstract":"In the last decades, the majority of the existing infrastructure heritage is approaching the end of its nominal design life mainly due to aging, deterioration, and degradation phenomena, threatening the safety levels of these strategic routes of communications. For civil engineers and researchers devoted to assessing and monitoring the structural health (SHM) of existing structures, the demand for innovative indirect non-destructive testing (NDT) methods aided with artificial intelligence (AI) is progressively spreading. In the present study, the authors analyzed the exertion of various deep learning models in order to increase the productivity of classifying ground penetrating radar (GPR) images for SHM purposes, especially focusing on road tunnel linings evaluations. Specifically, the authors presented a comparative study employing two convolutional models, i.e. the ResNet-50 and the EfficientNet-B0, and a recent transformer model, i.e. the Vision Transformer (ViT). Precisely, the authors evaluated the effects of training the models with or without pre-processed data through the bi-dimensional Fourier transform. Despite the theoretical advantages envisaged by adopting this kind of pre-processing technique on GPR images, the best classification performances have been still manifested by the classifiers trained without the Fourier pre-processing.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44766333","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Achieving Lifelong Open-ended Learning Autonomy (LOLA) is a key challenge in the field of robotics to advance to a new level of intelligent response. Robots should be capable of discovering goals and learn skills in specific domains that permit achieving the general objectives the designer establishes for them. In addition, robots should reuse previously learnt knowledge in different domains to facilitate learning and adaptation in new ones. To this end, cognitive architectures have arisen which encompass different components to support LOLA. A key feature of these architectures is to implement a proper balance between deliberative and reactive processes that allows for efficient real time operation and knowledge acquisition, but this is still an open issue. First, objectives must be defined in a domain-independent representation that allows for the autonomous determination of domain-dependent goals. Second, as no explicit reward function is available, a method to determine expected utility must also be developed. Finally, policy learning may happen in an internal deliberative scale (dreaming), so it is necessary to provide an efficient way to infer relevant and reliable data for dreaming to be meaningful. The first two aspects have already been addressed in the realm of the e-MDB cognitive architecture. For the third one, this work proposes Perceptual Classes (P-nodes) as a metacognitive structure that permits generating relevant “dreamt” data points that allow creating “imagined” trajectories for deliberative policy learning in a very efficient way. The proposed structure has been tested by means of an experiment with a real robot in LOLA settings, where it has been shown how policy dreaming is possible in such a challenging realm.
{"title":"Using perceptual classes to dream policies in open-ended learning robotics","authors":"A. Romero, Blaž Meden, F. Bellas, R. Duro","doi":"10.3233/ica-230707","DOIUrl":"https://doi.org/10.3233/ica-230707","url":null,"abstract":"Achieving Lifelong Open-ended Learning Autonomy (LOLA) is a key challenge in the field of robotics to advance to a new level of intelligent response. Robots should be capable of discovering goals and learn skills in specific domains that permit achieving the general objectives the designer establishes for them. In addition, robots should reuse previously learnt knowledge in different domains to facilitate learning and adaptation in new ones. To this end, cognitive architectures have arisen which encompass different components to support LOLA. A key feature of these architectures is to implement a proper balance between deliberative and reactive processes that allows for efficient real time operation and knowledge acquisition, but this is still an open issue. First, objectives must be defined in a domain-independent representation that allows for the autonomous determination of domain-dependent goals. Second, as no explicit reward function is available, a method to determine expected utility must also be developed. Finally, policy learning may happen in an internal deliberative scale (dreaming), so it is necessary to provide an efficient way to infer relevant and reliable data for dreaming to be meaningful. The first two aspects have already been addressed in the realm of the e-MDB cognitive architecture. For the third one, this work proposes Perceptual Classes (P-nodes) as a metacognitive structure that permits generating relevant “dreamt” data points that allow creating “imagined” trajectories for deliberative policy learning in a very efficient way. The proposed structure has been tested by means of an experiment with a real robot in LOLA settings, where it has been shown how policy dreaming is possible in such a challenging realm.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73890468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jose D. Fernández, Jorge García-González, Rafaela Benítez-Rochel, Miguel A. Molina-Cabello, Gonzalo Ramos-Jiménez, Ezequiel López-Rubio
Video feeds from traffic cameras can be useful for many purposes, the most critical of which are related to monitoring road safety. Vehicle trajectory is a key element in dangerous behavior and traffic accidents. In this respect, it is crucial to detect those anomalous vehicle trajectories, that is, trajectories that depart from usual paths. In this work, a model is proposed to automatically address that by using video sequences from traffic cameras. The proposal detects vehicles frame by frame, tracks their trajectories across frames, estimates velocity vectors, and compares them to velocity vectors from other spatially adjacent trajectories. From the comparison of velocity vectors, trajectories that are very different (anomalous) from neighboring trajectories can be detected. In practical terms, this strategy can detect vehicles in wrong-way trajectories. Some components of the model are off-the-shelf, such as the detection provided by recent deep learning approaches; however, several different options are considered and analyzed for vehicle tracking. The performance of the system has been tested with a wide range of real and synthetic traffic videos.
{"title":"Automated detection of vehicles with anomalous trajectories in traffic surveillance videos","authors":"Jose D. Fernández, Jorge García-González, Rafaela Benítez-Rochel, Miguel A. Molina-Cabello, Gonzalo Ramos-Jiménez, Ezequiel López-Rubio","doi":"10.3233/ica-230706","DOIUrl":"https://doi.org/10.3233/ica-230706","url":null,"abstract":"Video feeds from traffic cameras can be useful for many purposes, the most critical of which are related to monitoring road safety. Vehicle trajectory is a key element in dangerous behavior and traffic accidents. In this respect, it is crucial to detect those anomalous vehicle trajectories, that is, trajectories that depart from usual paths. In this work, a model is proposed to automatically address that by using video sequences from traffic cameras. The proposal detects vehicles frame by frame, tracks their trajectories across frames, estimates velocity vectors, and compares them to velocity vectors from other spatially adjacent trajectories. From the comparison of velocity vectors, trajectories that are very different (anomalous) from neighboring trajectories can be detected. In practical terms, this strategy can detect vehicles in wrong-way trajectories. Some components of the model are off-the-shelf, such as the detection provided by recent deep learning approaches; however, several different options are considered and analyzed for vehicle tracking. The performance of the system has been tested with a wide range of real and synthetic traffic videos.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75758804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. D. Fernández-Rodríguez, E. Palomo, J. M. Ortiz-de-Lazcano-Lobato, G. Ramos-Jiménez, Ezequiel López-Rubio
The dilemma between stability and plasticity is crucial in machine learning, especially when non-stationary input distributions are considered. This issue can be addressed by continual learning in order to alleviate catastrophic forgetting. This strategy has been previously proposed for supervised and reinforcement learning models. However, little attention has been devoted to unsupervised learning. This work presents a dynamic learning rate framework for unsupervised neural networks that can handle non-stationary distributions. In order for the model to adapt to the input as it changes its characteristics, a varying learning rate that does not merely depend on the training step but on the reconstruction error has been proposed. In the experiments, different configurations for classical competitive neural networks, self-organizing maps and growing neural gas with either per-neuron or per-network dynamic learning rate have been tested. Experimental results on document clustering tasks demonstrate the suitability of the proposal for real-world problems.
{"title":"Dynamic learning rates for continual unsupervised learning","authors":"J. D. Fernández-Rodríguez, E. Palomo, J. M. Ortiz-de-Lazcano-Lobato, G. Ramos-Jiménez, Ezequiel López-Rubio","doi":"10.3233/ica-230701","DOIUrl":"https://doi.org/10.3233/ica-230701","url":null,"abstract":"The dilemma between stability and plasticity is crucial in machine learning, especially when non-stationary input distributions are considered. This issue can be addressed by continual learning in order to alleviate catastrophic forgetting. This strategy has been previously proposed for supervised and reinforcement learning models. However, little attention has been devoted to unsupervised learning. This work presents a dynamic learning rate framework for unsupervised neural networks that can handle non-stationary distributions. In order for the model to adapt to the input as it changes its characteristics, a varying learning rate that does not merely depend on the training step but on the reconstruction error has been proposed. In the experiments, different configurations for classical competitive neural networks, self-organizing maps and growing neural gas with either per-neuron or per-network dynamic learning rate have been tested. Experimental results on document clustering tasks demonstrate the suitability of the proposal for real-world problems.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84622661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Robust and reliable 3D multi-object tracking (MOT) is essential for autonomous driving in crowded urban road scenes. In those scenarios, accurate data association between tracked objects and incoming new detections is crucial. This paper presents a tracking system based on the Kalman filter that uses a deep learning approach to the association problem. The proposed architecture consists of three neural networks. First, a convolutional LSTM network extracts spatiotemporal features from a sequence of detections of the same track. Then, a Siamese network calculates the degree of similarity between all tracks and the new detections found at each new frame. Finally, a recurrent LSTM network is used to extract 3D and bounding box information. This model follows the tracking-by-detection paradigm and has been trained with track sequences to be able to handle missed observations and to reduce identity switches. A validation test was carried out on the Argoverse dataset to validate the performance of the proposed system. The developed deep learning approach could improve current multi-object tracking systems based on classic algorithms like the Kalman filter.
{"title":"An improved deep learning architecture for multi-object tracking systems","authors":"Jesús Urdiales, David Martín, J. M. Armingol","doi":"10.3233/ica-230702","DOIUrl":"https://doi.org/10.3233/ica-230702","url":null,"abstract":"Robust and reliable 3D multi-object tracking (MOT) is essential for autonomous driving in crowded urban road scenes. In those scenarios, accurate data association between tracked objects and incoming new detections is crucial. This paper presents a tracking system based on the Kalman filter that uses a deep learning approach to the association problem. The proposed architecture consists of three neural networks. First, a convolutional LSTM network extracts spatiotemporal features from a sequence of detections of the same track. Then, a Siamese network calculates the degree of similarity between all tracks and the new detections found at each new frame. Finally, a recurrent LSTM network is used to extract 3D and bounding box information. This model follows the tracking-by-detection paradigm and has been trained with track sequences to be able to handle missed observations and to reduce identity switches. A validation test was carried out on the Argoverse dataset to validate the performance of the proposed system. The developed deep learning approach could improve current multi-object tracking systems based on classic algorithms like the Kalman filter.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82045396","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Canito, Armando Nobre, J. Neves, J. Corchado, G. Marreiros
In this paper, we present an architecture for time-constrained ontology evolution comprised of two tools: the J2OIM (JSON to Ontology Instance Mapper), which uses JavaScript Object Notation (JSON) objects to populate an ontology, and TICO (Time Constrained instance-guided Ontology evolution), which analyses streams or batches of instances as they are generated and attempts to identify potential changes to their definitions that may trigger evolutionary processes. These tools help compensate for identified gaps in literature in instance mapping and modular versioning. The case-study for these tools involves a predictive maintenance (PdM) scenario in which near real-time data sensor enriched by contextual data is continuously transformed into ontology individuals that trigger ontology evolution mechanisms. Results show it is possible to use the instance mapping mechanisms in an incremental fashion while assuring no duplicates are generated and the aggregation of similar information from distinct data points into intervals. Furthermore, they show how the ontology evolution processes effectively detect variations in ontology individuals, generating and updating existing concepts and roles.
{"title":"Using sensor data to detect time-constraints in ontology evolution","authors":"A. Canito, Armando Nobre, J. Neves, J. Corchado, G. Marreiros","doi":"10.3233/ica-230703","DOIUrl":"https://doi.org/10.3233/ica-230703","url":null,"abstract":"In this paper, we present an architecture for time-constrained ontology evolution comprised of two tools: the J2OIM (JSON to Ontology Instance Mapper), which uses JavaScript Object Notation (JSON) objects to populate an ontology, and TICO (Time Constrained instance-guided Ontology evolution), which analyses streams or batches of instances as they are generated and attempts to identify potential changes to their definitions that may trigger evolutionary processes. These tools help compensate for identified gaps in literature in instance mapping and modular versioning. The case-study for these tools involves a predictive maintenance (PdM) scenario in which near real-time data sensor enriched by contextual data is continuously transformed into ontology individuals that trigger ontology evolution mechanisms. Results show it is possible to use the instance mapping mechanisms in an incremental fashion while assuring no duplicates are generated and the aggregation of similar information from distinct data points into intervals. Furthermore, they show how the ontology evolution processes effectively detect variations in ontology individuals, generating and updating existing concepts and roles.","PeriodicalId":50358,"journal":{"name":"Integrated Computer-Aided Engineering","volume":null,"pages":null},"PeriodicalIF":6.5,"publicationDate":"2023-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80679541","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","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：