Pub Date : 2019-12-01DOI: 10.1109/ICMLA.2019.00041
S. Nageshrao, Bruno Costa, Dimitar Filev
In many industrial applications, one of the major bottlenecks in using advanced learning-based methods (such as reinforcement learning) for controls is the lack of interpretability of the trained agent. In this paper, we present a methodology for translating a trained reinforcement learning agent into a set of simple and easy to interpret if-then rules by using the proven universal approximation property of the rules with fuzzy predicates. Proposed methodology combines the optimality of reinforcement learning with interpretability of the theory of approximate reasoning, thus making reinforcement learning-based solutions more accessible to industrial practitioners. The framework presented in this paper has the potential to help address the fundamental problem in widespread adoption of reinforcement learning in industrial applications.
{"title":"Interpretable Approximation of a Deep Reinforcement Learning Agent as a Set of If-Then Rules","authors":"S. Nageshrao, Bruno Costa, Dimitar Filev","doi":"10.1109/ICMLA.2019.00041","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00041","url":null,"abstract":"In many industrial applications, one of the major bottlenecks in using advanced learning-based methods (such as reinforcement learning) for controls is the lack of interpretability of the trained agent. In this paper, we present a methodology for translating a trained reinforcement learning agent into a set of simple and easy to interpret if-then rules by using the proven universal approximation property of the rules with fuzzy predicates. Proposed methodology combines the optimality of reinforcement learning with interpretability of the theory of approximate reasoning, thus making reinforcement learning-based solutions more accessible to industrial practitioners. The framework presented in this paper has the potential to help address the fundamental problem in widespread adoption of reinforcement learning in industrial applications.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134009517","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00316
M. Boger, A. Laverghetta, Nikolai Fetisov, John Licato
Analogical reasoning, it has been argued, fundamentally underlies many cognitive processes and is an important marker of developmental cognition. This connection suggests that the clever use of analogical reasoning tasks can improve cognitive performance in specific ways, thus leading to clear educational applications, as recent psychological work has confirmed. However, currently there are no known methods to either solve or generate analogical word problems, at least to a degree of reliability that would be necessary before such educational applications are possible. To address these concerns we present work to both solve and generate analogy word problems: First, given an analogy word problem, our algorithm performs a parallel random walk through the semantic network ConceptNet to limit the number of choices that are then considered by a vector embedding. We achieve an improvement in accuracy beyond existing state-of-the-art. Second, we explore a method for automatically generating explainable n-step analogy word problems, and analyze the results.
{"title":"Generating Near and Far Analogies for Educational Applications: Progress and Challenges","authors":"M. Boger, A. Laverghetta, Nikolai Fetisov, John Licato","doi":"10.1109/ICMLA.2019.00316","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00316","url":null,"abstract":"Analogical reasoning, it has been argued, fundamentally underlies many cognitive processes and is an important marker of developmental cognition. This connection suggests that the clever use of analogical reasoning tasks can improve cognitive performance in specific ways, thus leading to clear educational applications, as recent psychological work has confirmed. However, currently there are no known methods to either solve or generate analogical word problems, at least to a degree of reliability that would be necessary before such educational applications are possible. To address these concerns we present work to both solve and generate analogy word problems: First, given an analogy word problem, our algorithm performs a parallel random walk through the semantic network ConceptNet to limit the number of choices that are then considered by a vector embedding. We achieve an improvement in accuracy beyond existing state-of-the-art. Second, we explore a method for automatically generating explainable n-step analogy word problems, and analyze the results.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133529007","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00272
Marc Wenninger, Dominik Stecher, Jochen Schmidt
Generating a more detailed understanding of domestic electricity demand is a major topic for energy suppliers and householders in times of climate change. Over the years there have been many studies on consumption feedback systems to inform householders, disaggregation algorithms for Non-Intrusive-Load-Monitoring (NILM), Real-Time-Pricing (RTP) to promote supply aware behavior through monetary incentives and appliance usage prediction algorithms. While these studies are vital steps towards energy awareness, one of the most fundamental challenges has not yet been tackled: Automated detection of start and stop of usage cycles of household appliances. We argue that most research efforts in this area will benefit from a reliable segmentation method to provide accurate usage information. We propose a SVM-based segmentation method for home appliances such as dishwashers and washing machines. The method is evaluated using manually annotated electricity measurements of five different appliances recorded over two years in multiple households.
{"title":"SVM-Based Segmentation of Home Appliance Energy Measurements","authors":"Marc Wenninger, Dominik Stecher, Jochen Schmidt","doi":"10.1109/ICMLA.2019.00272","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00272","url":null,"abstract":"Generating a more detailed understanding of domestic electricity demand is a major topic for energy suppliers and householders in times of climate change. Over the years there have been many studies on consumption feedback systems to inform householders, disaggregation algorithms for Non-Intrusive-Load-Monitoring (NILM), Real-Time-Pricing (RTP) to promote supply aware behavior through monetary incentives and appliance usage prediction algorithms. While these studies are vital steps towards energy awareness, one of the most fundamental challenges has not yet been tackled: Automated detection of start and stop of usage cycles of household appliances. We argue that most research efforts in this area will benefit from a reliable segmentation method to provide accurate usage information. We propose a SVM-based segmentation method for home appliances such as dishwashers and washing machines. The method is evaluated using manually annotated electricity measurements of five different appliances recorded over two years in multiple households.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130279786","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00019
Dan Halbersberg, B. Lerner
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease, lasting from the day of onset until death. Factors such as the progression rate and pattern of the disease vary greatly among patients, making it difficult to achieve accurate predictions about ALS. To accurately predict ALS disease state and deterioration, we propose a novel approach that combines: a) sequence clustering based on dynamic time warping for separation among patients with diverse ALS deterioration patterns, b) sequential pattern mining for discovery of deterioration changes that patients of the same type may have in common, and c) deterioration-based patient next-state prediction. Using a clinical dataset, we demonstrate the advantage of the proposed approach in terms of classification accuracy and deterioration detection compared to other classification methods and temporal models such as long short-term memory.
{"title":"Temporal Modeling of Deterioration Patterns and Clustering for Disease Prediction of ALS Patients","authors":"Dan Halbersberg, B. Lerner","doi":"10.1109/ICMLA.2019.00019","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00019","url":null,"abstract":"Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease, lasting from the day of onset until death. Factors such as the progression rate and pattern of the disease vary greatly among patients, making it difficult to achieve accurate predictions about ALS. To accurately predict ALS disease state and deterioration, we propose a novel approach that combines: a) sequence clustering based on dynamic time warping for separation among patients with diverse ALS deterioration patterns, b) sequential pattern mining for discovery of deterioration changes that patients of the same type may have in common, and c) deterioration-based patient next-state prediction. Using a clinical dataset, we demonstrate the advantage of the proposed approach in terms of classification accuracy and deterioration detection compared to other classification methods and temporal models such as long short-term memory.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130330356","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00089
A. Muthukumar, Ramakrishnan Durairajan
To understand the delay characteristics of the Internet, a myriad of measurement tools and techniques are proposed by the researchers in academia and industry. Datasets from such measurement tools are curated to facilitate analyses at a later time. Despite the benefits of these tools and datasets, the systematic interpretation of measurements in the face of measurement noise. Unfortunately, state-of-the-art denoising techniques are labor-intensive and ineffective. To tackle this problem, we develop NoMoNoise, an open-source framework for denoising latency measurements by leveraging the recent advancements in weak-supervised learning. NoMoNoise can generate measurement noise labels that could be integrated into the inference and control logic to remove and/or repair noisy measurements in an automated and rapid fashion. We evaluate the efficacy of NoMoNoise in a lab-based setting and a real-world setting by applying it on CAIDA's Ark dataset and show that NoMoNoise can remove noisy measurements effectively with high accuracy.
{"title":"Denoising Internet Delay Measurements using Weak Supervision","authors":"A. Muthukumar, Ramakrishnan Durairajan","doi":"10.1109/ICMLA.2019.00089","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00089","url":null,"abstract":"To understand the delay characteristics of the Internet, a myriad of measurement tools and techniques are proposed by the researchers in academia and industry. Datasets from such measurement tools are curated to facilitate analyses at a later time. Despite the benefits of these tools and datasets, the systematic interpretation of measurements in the face of measurement noise. Unfortunately, state-of-the-art denoising techniques are labor-intensive and ineffective. To tackle this problem, we develop NoMoNoise, an open-source framework for denoising latency measurements by leveraging the recent advancements in weak-supervised learning. NoMoNoise can generate measurement noise labels that could be integrated into the inference and control logic to remove and/or repair noisy measurements in an automated and rapid fashion. We evaluate the efficacy of NoMoNoise in a lab-based setting and a real-world setting by applying it on CAIDA's Ark dataset and show that NoMoNoise can remove noisy measurements effectively with high accuracy.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125649665","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00312
O. Iloanusi, C. Mbah
High gender classification accuracies have been recorded with high-resolution faces under controlled conditions. However, real-life scenarios are faced with challenges not limited to high pose variations in subjects, poor visibility, occlusion, and distance from camera. These have led to the current trend in estimating gender from full body images, notwithstanding the challenges posed by partial body images in a typical life scenario. We demonstrate that there are certain sections in a body image, the face, upper or lower body that are useful for recognition at near or far distances. Given the challenges of body captured at far distance or partially showing body in a photo, we therefore propose a combination of three classifiers for gender estimation from face; upper and full body from single-shot image. Our results in far compared to near distance images suggest that gender is best estimated from a hybrid of face; upper and full body images under challenging conditions.
{"title":"Gender Estimation from a Hybrid of Face, Upper and Full Body Images at Varying Body Poses","authors":"O. Iloanusi, C. Mbah","doi":"10.1109/ICMLA.2019.00312","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00312","url":null,"abstract":"High gender classification accuracies have been recorded with high-resolution faces under controlled conditions. However, real-life scenarios are faced with challenges not limited to high pose variations in subjects, poor visibility, occlusion, and distance from camera. These have led to the current trend in estimating gender from full body images, notwithstanding the challenges posed by partial body images in a typical life scenario. We demonstrate that there are certain sections in a body image, the face, upper or lower body that are useful for recognition at near or far distances. Given the challenges of body captured at far distance or partially showing body in a photo, we therefore propose a combination of three classifiers for gender estimation from face; upper and full body from single-shot image. Our results in far compared to near distance images suggest that gender is best estimated from a hybrid of face; upper and full body images under challenging conditions.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132396798","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00232
Hossein Hassani, Maryam Farajzadeh-Zanjani, R. Razavi-Far, M. Saif, V. Palade
There has been a growing interest in using smart grids due to their capability in delivering automated and distributed energy level to the consumption units. However, in order to guarantee the safe and reliable delivery of the high-quality power from the generation units to the consumers, smart grids need to be equipped with diagnostic systems. This paper presents an efficient data-driven scheme for diagnosing faults in smart grids. In order to reduce the computational burden and monitor the state of the system with a lower number of smart meters, a method based on the affinity propagation clustering algorithm is suggested for the placement of meters, that makes use of the graph-based representation of the system. The collected voltage data measurements from the installed meters are then decomposed by matching pursuit decomposition in order to generate informative features. Extracted features are then used to train a convolutional neural network, and the constructed deep learning model is then tested using unseen samples of normal and faulty conditions. Simulation results based on the IEEE 39–Bus System demonstrate the effectiveness of the proposed data-driven fault diagnostic system.
{"title":"Design of a Cost-Effective Deep Convolutional Neural Network–Based Scheme for Diagnosing Faults in Smart Grids","authors":"Hossein Hassani, Maryam Farajzadeh-Zanjani, R. Razavi-Far, M. Saif, V. Palade","doi":"10.1109/ICMLA.2019.00232","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00232","url":null,"abstract":"There has been a growing interest in using smart grids due to their capability in delivering automated and distributed energy level to the consumption units. However, in order to guarantee the safe and reliable delivery of the high-quality power from the generation units to the consumers, smart grids need to be equipped with diagnostic systems. This paper presents an efficient data-driven scheme for diagnosing faults in smart grids. In order to reduce the computational burden and monitor the state of the system with a lower number of smart meters, a method based on the affinity propagation clustering algorithm is suggested for the placement of meters, that makes use of the graph-based representation of the system. The collected voltage data measurements from the installed meters are then decomposed by matching pursuit decomposition in order to generate informative features. Extracted features are then used to train a convolutional neural network, and the constructed deep learning model is then tested using unseen samples of normal and faulty conditions. Simulation results based on the IEEE 39–Bus System demonstrate the effectiveness of the proposed data-driven fault diagnostic system.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"257 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134103142","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00151
Truc The Nguyen, Alexander Fuchs, F. Pernkopf
We propose a heterogeneous system of Deep Mixture of Experts (DMoEs) models using different Convolutional Neural Networks (CNNs) for acoustic scene classification (ASC). Each DMoEs module is a mixture of different parallel CNN structures weighted by a gating network. All CNNs use the same input data. The CNN architectures play the role of experts extracting a variety of features. The experts are pre-trained, and kept fixed (frozen) for the DMoEs model. The DMoEs is post-trained by optimizing weights of the gating network, which estimates the contribution of the experts in the mixture. In order to enhance the performance, we use an ensemble of three DMoEs modules each with different pairs of inputs and individual CNN models. The input pairs are spectrogram combinations of binaural audio and mono audio as well as their pre-processed variations using harmonic-percussive source separation (HPSS) and nearest neighbor filters (NNFs). The classification result of the proposed system is 72.1% improving the baseline by around 12% (absolute) on the development data of DCASE 2018 challenge task 1A.
{"title":"Acoustic Scene Classification Using Deep Mixtures of Pre-trained Convolutional Neural Networks","authors":"Truc The Nguyen, Alexander Fuchs, F. Pernkopf","doi":"10.1109/ICMLA.2019.00151","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00151","url":null,"abstract":"We propose a heterogeneous system of Deep Mixture of Experts (DMoEs) models using different Convolutional Neural Networks (CNNs) for acoustic scene classification (ASC). Each DMoEs module is a mixture of different parallel CNN structures weighted by a gating network. All CNNs use the same input data. The CNN architectures play the role of experts extracting a variety of features. The experts are pre-trained, and kept fixed (frozen) for the DMoEs model. The DMoEs is post-trained by optimizing weights of the gating network, which estimates the contribution of the experts in the mixture. In order to enhance the performance, we use an ensemble of three DMoEs modules each with different pairs of inputs and individual CNN models. The input pairs are spectrogram combinations of binaural audio and mono audio as well as their pre-processed variations using harmonic-percussive source separation (HPSS) and nearest neighbor filters (NNFs). The classification result of the proposed system is 72.1% improving the baseline by around 12% (absolute) on the development data of DCASE 2018 challenge task 1A.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134398568","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00235
Xueqin Pang, C. Forrest, F. Lê-Scherban, A. Masino
Obesity, as an independent risk factor for increased morbidity and mortality throughout the lifecycle, is a major health issue in the United States. Pediatric obesity is a strong risk factor for adult obesity, as it tends to be stable and tracks into adulthood. Therefore, prevention of childhood obesity is urgently required for reduction in obesity prevalence and obesity related comorbidities. In this paper, the general pediatric obesity development pattern and the onset time period of early childhood obesity was identified via analysis of approximately 11 million pediatric clinical encounters of 860,510 unique individuals. XGBoost model was developed to predict at age 2 years if individuals would develop obesity in early childhood. The model is generalized to both males and females, and achieved an AUC of 81% (± 0.1%). Obesity associated risk factors were further analyzed via interpretation of the XGBoost model predictions. Besides known predictive factors such as weight, height, race, and ethnicity, new factors such as body temperature and respiratory rate were also identified. As body temperature and respiratory rate are related to human metabolism, novel physiologic mechanisms that cause these associations might be discovered in future research. We decomposed model recall to different age ranges when obesity incidence occurred. The model recall for individuals with obesity incidence between 24–36 months was 97.63%, while recall for obesity incidence between 72–84 months was 48.96%, suggesting obesity is less predictable further in the future. Since obesity is largely affected by evolving factors such as life style, diet, and living environment, it is possible that obesity prevention may be achieved via changes in adjustable factors.
{"title":"Understanding Early Childhood Obesity via Interpretation of Machine Learning Model Predictions","authors":"Xueqin Pang, C. Forrest, F. Lê-Scherban, A. Masino","doi":"10.1109/ICMLA.2019.00235","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00235","url":null,"abstract":"Obesity, as an independent risk factor for increased morbidity and mortality throughout the lifecycle, is a major health issue in the United States. Pediatric obesity is a strong risk factor for adult obesity, as it tends to be stable and tracks into adulthood. Therefore, prevention of childhood obesity is urgently required for reduction in obesity prevalence and obesity related comorbidities. In this paper, the general pediatric obesity development pattern and the onset time period of early childhood obesity was identified via analysis of approximately 11 million pediatric clinical encounters of 860,510 unique individuals. XGBoost model was developed to predict at age 2 years if individuals would develop obesity in early childhood. The model is generalized to both males and females, and achieved an AUC of 81% (± 0.1%). Obesity associated risk factors were further analyzed via interpretation of the XGBoost model predictions. Besides known predictive factors such as weight, height, race, and ethnicity, new factors such as body temperature and respiratory rate were also identified. As body temperature and respiratory rate are related to human metabolism, novel physiologic mechanisms that cause these associations might be discovered in future research. We decomposed model recall to different age ranges when obesity incidence occurred. The model recall for individuals with obesity incidence between 24–36 months was 97.63%, while recall for obesity incidence between 72–84 months was 48.96%, suggesting obesity is less predictable further in the future. Since obesity is largely affected by evolving factors such as life style, diet, and living environment, it is possible that obesity prevention may be achieved via changes in adjustable factors.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131694589","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 : 2019-12-01DOI: 10.1109/ICMLA.2019.00062
Laura Dörr, Felix Brandt, Anne Meyer, Martin Pouls
Supervised deep learning has become the state of the art method for object detection and is used in many application areas such as autonomous driving, manufacturing industries or security systems. The acquisition of annotated data sets for the training of neural networks is highly time-consuming and error-prone. Thus, the supervised training of such object detection models is not feasible in some cases. This holds for the task of logistics transport label detection, as this use-case stands out by requiring highly specialized, quickly adapting models whilst allowing for little manual efforts in the data preparation and training process. We propose an easy training data generation method enabling the fully automated training of specialized models for the task of logistics transport label detection. For data synthesis, we stitch instances of the transport labels to be detected into background images whilst using image degradation and augmentation methods. We evaluate the employment of both use-case-specific, carefully selected background images and randomly selected real-world background images. Further, we compare two different data generation approaches: one generating realistically looking images and a simpler one making do without any manual image annotation. We examine and evaluate the introduced method on a new and publicly available example data set relevant for logistics transport label detection. We show that accurate models can be trained exclusively on synthetic training data and we compare their performance to models trained on real, manually annotated images.
{"title":"Lean Training Data Generation for Planar Object Detection Models in Unsteady Logistics Contexts","authors":"Laura Dörr, Felix Brandt, Anne Meyer, Martin Pouls","doi":"10.1109/ICMLA.2019.00062","DOIUrl":"https://doi.org/10.1109/ICMLA.2019.00062","url":null,"abstract":"Supervised deep learning has become the state of the art method for object detection and is used in many application areas such as autonomous driving, manufacturing industries or security systems. The acquisition of annotated data sets for the training of neural networks is highly time-consuming and error-prone. Thus, the supervised training of such object detection models is not feasible in some cases. This holds for the task of logistics transport label detection, as this use-case stands out by requiring highly specialized, quickly adapting models whilst allowing for little manual efforts in the data preparation and training process. We propose an easy training data generation method enabling the fully automated training of specialized models for the task of logistics transport label detection. For data synthesis, we stitch instances of the transport labels to be detected into background images whilst using image degradation and augmentation methods. We evaluate the employment of both use-case-specific, carefully selected background images and randomly selected real-world background images. Further, we compare two different data generation approaches: one generating realistically looking images and a simpler one making do without any manual image annotation. We examine and evaluate the introduced method on a new and publicly available example data set relevant for logistics transport label detection. We show that accurate models can be trained exclusively on synthetic training data and we compare their performance to models trained on real, manually annotated images.","PeriodicalId":436714,"journal":{"name":"2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133253828","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：