Frequently, high-dimensional features are used to represent data to be classified. This paper pro... more Frequently, high-dimensional features are used to represent data to be classified. This paper proposes a new approach to learn interpretable classification models from such high-dimensional data representation. To this end, we extend a popular prototype-based classification algorithm, the matrix learning vector quantization, to incorporate an enhanced feature selection objective via \(L_1\)-regularization. In contrast to previous work, we propose a framework that directly optimizes this objective using the alternating direction method of multipliers (ADMM) and manifold optimization. We evaluate our method on synthetic data and on real data for speech-based emotion recognition. Particularly, we show that our method achieves state-of-the-art results on the Berlin Database of Emotional speech and show its abilities to select relevant dimensions from the eGeMAPS set of audio features.
This paper presents a fast but yet simple solution to create areas on metrical occupancy grid map... more This paper presents a fast but yet simple solution to create areas on metrical occupancy grid maps which can easily be converted to topological maps. Those maps with their areas provide an understandable “non-expert” view on a robots environment and allow semantic labeling. This serves as a foundation for supporting navigation tasks like path planning, localization and human-machine-interaction which are not in the scope of this paper.
Fig. 1. One of our robots. The research goal of the Artificial Intelligence Lab at the University... more Fig. 1. One of our robots. The research goal of the Artificial Intelligence Lab at the University of Applied Sciences Dresden is the development of intelligent and interactive mobile service and assistance systems. For the application of such systems, different scenarios are imaginable, like guidance and information systems within public institutions (museums, airports, train stations, etc.), assistance or supporting systems in markets (do-it-yourself store, shopping malls,etc.), systems to support the elderly in home environments, and systems for industrial tasks (mobile measuring and surveillance systems). For the current research work the robot, which is depicted in Fig. 1, is used as a guidance system in a museum in the city of Dresden (Technische Sammlungen Dresden). The research platform is a commercially available SCITOS G5 equipped with two laser range finders, an omnidirectional camera, a depth camera, a ring of sonar sensors, and a head with two eyes that has no further se...
Current robotic software frameworks lack a mean to aid in the generation, validation and presenta... more Current robotic software frameworks lack a mean to aid in the generation, validation and presentation of high quality content for user interaction. This paper introduces a new approach to extend a basic robotic software framework with a layer for content management. This layer has capabilities for controlling the content presentation subsystems already integrated. We introduce abstract dialog acts as a centerpiece for creating and handling robot behavior including user interactions. A simple file format is used to edit the dialog act structure and it allows the delegation of the dialog creation to domain experts within the desired field. We demonstrate that the creation of different sets of dialog acts allows the implementation of completely different use cases without requiring any changes to existing software components.
International Journal of Computers and Applications
Abstract With the project "FIPS" (Ferngesteuertes Internet Praktikum Schaltsysteme—in E... more Abstract With the project "FIPS" (Ferngesteuertes Internet Praktikum Schaltsysteme—in English: remote controlled laboratory for the course “Digital Logic Design”) this contribution introduces a concept of a Web-Based Laboratory with several experiments offering controlling, observing, and analysing features via internet. All single steps of the laboratory process—starting with the design of control algorithms up to the web-based validation—will be presented on the basis of the existing laboratory experiment “Travelling Trolley". Derived from that, the required system architecture of the project “FIPS" will be shown and different kinds of realization will be compared regarding the execution of the designed control algorithm.
This paper presents an approach to multi-object image segmentation based on object motion using M... more This paper presents an approach to multi-object image segmentation based on object motion using Markov Random Fields 1. To support the information gained from motion and to achieve robustness, several additional visual cues extracted from the image data are integrated. Depth information gained from stereo disparity is included to maintain segmentation in case an segmented object stops moving. Motion is estimated with a correspondence matching scheme. The approach differs from regular optical flow in the way that rich matching results are used for segmentation rather than only the best matches. The representation of segmented regions is realized implicitly as labeling on a 2D lattice. Motion segmentation is a key to many modern image processing applications. In video compression algorithms, the analysis of motion and regions with coherent motion helps to drastically reduce the amount of information that has to be stored and transmitted for each frame [11]. Motion segmentation and motion understanding, for example, plays an essential role in detecting and/or avoiding obstacles in vehicles or with a mobile robot. The rest of this paper is organized as follows: while Sect. 2 provides a short overview of the work done in the field of image segmentation and relations to our approach, Sect. 3 describes the architecture of the proposed system. Its evaluation is presented in Sect. 4. Finally, the paper concludes with Sect. 5. 2 State of the Art Recent research in the area of motion segmentation focuses on feature-based motion estimation, approaches using level set methods [4], and on multi-cue segmentation on Markov Random Fields [6]. Optimization on Markov Random Fields is an established approach to segmentation [9]. It permits to let motion estimation be part of the segmentation process in one single optimization framework. When Markov Random Fields are used in combination with optimization methods like iterated conditional modes (ICM) [2], the initialization of the field configuration and the label parameters has a strong influence on the time needed for convergence as well as on the quality of the result. In multi-frame applications, results from the last frame can be used for initialization [3]. Our approach uses motion estimates that fulfill certain quality criteria, as well as results from the previous frame.
and prediction. Consider the simple task for a robot of grasping an object which is handed over b... more and prediction. Consider the simple task for a robot of grasping an object which is handed over by the human interaction partner. To avoid a purely reactive behaviour, which might lead to ‘mechanical’ movements of the robots, it is necessary to predict the further movement of the human’s hand. In [2] an interesting concept for a decomposition task is presented. Like playing a piano a basis alphabet { the dierent notes { are superimposed to reconstruct the observation (the piece of music). Regarding only the information, when a base primitive was active, gives rise to an instance of the so called ’piano model’ which is a very low-dimensional and sparse representation and which can be exploited for further processing. While the so-called piano model relies on a set of given basis primitives, our approach is able to learn these primitives from the training data. We use [3], a blind source separation approach in concept similar to PCA and ICA. The system of basis vectors which is generated by the NMF is not orthogonal. This is very useful for motion trajectories, since one basis primitive is allowed to share a common part of its trajectory with other primitives and to specialize later.
This paper\'s intention is to adapt prediction algorithms well known in the field of time ser... more This paper\'s intention is to adapt prediction algorithms well known in the field of time series analysis to problems being faced it the field of mobile robotics and Human-Computer-Interation (HRI). The idea is to predict movement data by understanding it as time series. The prediction takes place with o black box model, which means that no further knowledge on motion dynamics is used then the post of the trajectory itself:This means, the suggested approaches are able to adapt yo different situations. Several state-of-the-art algorithms such as Local Modeling, Cluster Weighted Modeling, Echo State Networks and Autoregressive Models are evaluated and compared. For experiments, real movement trajectories of a human are used. Since mobile robots highly depend on real-time application, computing time is also considered. Experiments show that Echo State Networks and Local Models show impressive results for long term motion prediction with a prediction horizon of up to eight seconds.
Advances in Intelligent Systems and Computing, 2014
This paper aims at an approach for labeling places within a grid cell environment. For that we pr... more This paper aims at an approach for labeling places within a grid cell environment. For that we propose a method that is based on non-negative matrix factorization (NMF) to extract environment specific features from a given occupancy grid map. NMF also computes a description about where on the map these features need to be applied. We use this description after certain pre-processing steps as an input for generalized learning vector quantization (GLVQ) to achieve the classification or labeling of the grid cells. Our approach is evaluated on a standard data set from University of Freiburg, showing very promising results.
Der Beitrag stellt mit dem Projekt „FIPS“ das Konzept eines Web-basierten Praktikums mit verschie... more Der Beitrag stellt mit dem Projekt „FIPS“ das Konzept eines Web-basierten Praktikums mit verschiedenen Experimenten vor, die uber das Internet gesteuert, beobachtet und ausgewertet werden konnen. Dabei sollen neben den verschiedenen Realisierungsmoglichkeiten auch die einzelnen Schritte der Versuchsdurchfuhrung — angefangen vom Entwurf des Steuerungsalgorithmus bis hin zur web-gesteuerten Erprobung am realen Modell — anhand des Praktikumsversuches „2D-Laufkatzensteuerung“ vorgestellt werden. Darauf aufbauend werden die im Rahmen des Projektes „FIPS“ entwickelte Systemarchitektur vorgestellt sowie verschiedene Realisierungsmoglichkeiten bezuglich der Abarbeitung des entworfenen Steuerungsalgorithmus aufgezeigt und miteinander verglichen.
ABSTRACT We propose a system for enhancing a human-robot interaction system. The goal is to embra... more ABSTRACT We propose a system for enhancing a human-robot interaction system. The goal is to embrace surrounding persons into a conversation with the robot by means of establishing or keeping eye contact. For this, a simple and hence computationally efficient people tracking algorithm has been developed, which in itself is an enhancement of existing approaches. As the foundation of our approach we use a vanilla Kalman filter that explicitly models the velocity in the state space, while the observations are transformed into a unified global coordinate system. We evaluate different integration strategies for multiple sensor cues. Furthermore, we extend an algorithm for group detection to be able to recognize individuals.
2009 IEEE International Conference on Robotics and Automation, 2009
Recent advances in the field of humanoid robotics increase the complexity of the tasks that such ... more Recent advances in the field of humanoid robotics increase the complexity of the tasks that such robots can perform. This makes it increasingly difficult and inconvenient to program these tasks manually. Furthermore, humanoid robots, in contrast to industrial robots, should in the distant future behave within a social environment. Therefore, it must be possible to extend the robot's abilities in an easy and natural way. To address these requirements, this work investigates the topic of imitation learning of motor skills. The focus lies on providing a humanoid robot with the ability to learn new bimanual tasks through the observation of object trajectories. For this, an imitation learning framework is presented, which allows the robot to learn the important elements of an observed movement task by application of probabilistic encoding with Gaussian Mixture Models. The learned information is used to initialize an attractor-based movement generation algorithm that optimizes the reproduced movement towards the fulfillment of additional criteria, such as collision avoidance. Experiments performed with the humanoid robot ASIMO show that the proposed system is suitable for transferring information from a human demonstrator to the robot. These results provide a good starting point for more complex and interactive learning tasks.
ABSTRACT While he idea of tour guide robots has been addressed several times, applying a video-pr... more ABSTRACT While he idea of tour guide robots has been addressed several times, applying a video-projector based augmented reality component to this scenario is rather new. We show requirements of the localization system of the robot and how they can be fulfilled, as well as a basic system for projection correction and its suitability for this scenario.
Frequently, high-dimensional features are used to represent data to be classified. This paper pro... more Frequently, high-dimensional features are used to represent data to be classified. This paper proposes a new approach to learn interpretable classification models from such high-dimensional data representation. To this end, we extend a popular prototype-based classification algorithm, the matrix learning vector quantization, to incorporate an enhanced feature selection objective via \(L_1\)-regularization. In contrast to previous work, we propose a framework that directly optimizes this objective using the alternating direction method of multipliers (ADMM) and manifold optimization. We evaluate our method on synthetic data and on real data for speech-based emotion recognition. Particularly, we show that our method achieves state-of-the-art results on the Berlin Database of Emotional speech and show its abilities to select relevant dimensions from the eGeMAPS set of audio features.
This paper presents a fast but yet simple solution to create areas on metrical occupancy grid map... more This paper presents a fast but yet simple solution to create areas on metrical occupancy grid maps which can easily be converted to topological maps. Those maps with their areas provide an understandable “non-expert” view on a robots environment and allow semantic labeling. This serves as a foundation for supporting navigation tasks like path planning, localization and human-machine-interaction which are not in the scope of this paper.
Fig. 1. One of our robots. The research goal of the Artificial Intelligence Lab at the University... more Fig. 1. One of our robots. The research goal of the Artificial Intelligence Lab at the University of Applied Sciences Dresden is the development of intelligent and interactive mobile service and assistance systems. For the application of such systems, different scenarios are imaginable, like guidance and information systems within public institutions (museums, airports, train stations, etc.), assistance or supporting systems in markets (do-it-yourself store, shopping malls,etc.), systems to support the elderly in home environments, and systems for industrial tasks (mobile measuring and surveillance systems). For the current research work the robot, which is depicted in Fig. 1, is used as a guidance system in a museum in the city of Dresden (Technische Sammlungen Dresden). The research platform is a commercially available SCITOS G5 equipped with two laser range finders, an omnidirectional camera, a depth camera, a ring of sonar sensors, and a head with two eyes that has no further se...
Current robotic software frameworks lack a mean to aid in the generation, validation and presenta... more Current robotic software frameworks lack a mean to aid in the generation, validation and presentation of high quality content for user interaction. This paper introduces a new approach to extend a basic robotic software framework with a layer for content management. This layer has capabilities for controlling the content presentation subsystems already integrated. We introduce abstract dialog acts as a centerpiece for creating and handling robot behavior including user interactions. A simple file format is used to edit the dialog act structure and it allows the delegation of the dialog creation to domain experts within the desired field. We demonstrate that the creation of different sets of dialog acts allows the implementation of completely different use cases without requiring any changes to existing software components.
International Journal of Computers and Applications
Abstract With the project "FIPS" (Ferngesteuertes Internet Praktikum Schaltsysteme—in E... more Abstract With the project "FIPS" (Ferngesteuertes Internet Praktikum Schaltsysteme—in English: remote controlled laboratory for the course “Digital Logic Design”) this contribution introduces a concept of a Web-Based Laboratory with several experiments offering controlling, observing, and analysing features via internet. All single steps of the laboratory process—starting with the design of control algorithms up to the web-based validation—will be presented on the basis of the existing laboratory experiment “Travelling Trolley". Derived from that, the required system architecture of the project “FIPS" will be shown and different kinds of realization will be compared regarding the execution of the designed control algorithm.
This paper presents an approach to multi-object image segmentation based on object motion using M... more This paper presents an approach to multi-object image segmentation based on object motion using Markov Random Fields 1. To support the information gained from motion and to achieve robustness, several additional visual cues extracted from the image data are integrated. Depth information gained from stereo disparity is included to maintain segmentation in case an segmented object stops moving. Motion is estimated with a correspondence matching scheme. The approach differs from regular optical flow in the way that rich matching results are used for segmentation rather than only the best matches. The representation of segmented regions is realized implicitly as labeling on a 2D lattice. Motion segmentation is a key to many modern image processing applications. In video compression algorithms, the analysis of motion and regions with coherent motion helps to drastically reduce the amount of information that has to be stored and transmitted for each frame [11]. Motion segmentation and motion understanding, for example, plays an essential role in detecting and/or avoiding obstacles in vehicles or with a mobile robot. The rest of this paper is organized as follows: while Sect. 2 provides a short overview of the work done in the field of image segmentation and relations to our approach, Sect. 3 describes the architecture of the proposed system. Its evaluation is presented in Sect. 4. Finally, the paper concludes with Sect. 5. 2 State of the Art Recent research in the area of motion segmentation focuses on feature-based motion estimation, approaches using level set methods [4], and on multi-cue segmentation on Markov Random Fields [6]. Optimization on Markov Random Fields is an established approach to segmentation [9]. It permits to let motion estimation be part of the segmentation process in one single optimization framework. When Markov Random Fields are used in combination with optimization methods like iterated conditional modes (ICM) [2], the initialization of the field configuration and the label parameters has a strong influence on the time needed for convergence as well as on the quality of the result. In multi-frame applications, results from the last frame can be used for initialization [3]. Our approach uses motion estimates that fulfill certain quality criteria, as well as results from the previous frame.
and prediction. Consider the simple task for a robot of grasping an object which is handed over b... more and prediction. Consider the simple task for a robot of grasping an object which is handed over by the human interaction partner. To avoid a purely reactive behaviour, which might lead to ‘mechanical’ movements of the robots, it is necessary to predict the further movement of the human’s hand. In [2] an interesting concept for a decomposition task is presented. Like playing a piano a basis alphabet { the dierent notes { are superimposed to reconstruct the observation (the piece of music). Regarding only the information, when a base primitive was active, gives rise to an instance of the so called ’piano model’ which is a very low-dimensional and sparse representation and which can be exploited for further processing. While the so-called piano model relies on a set of given basis primitives, our approach is able to learn these primitives from the training data. We use [3], a blind source separation approach in concept similar to PCA and ICA. The system of basis vectors which is generated by the NMF is not orthogonal. This is very useful for motion trajectories, since one basis primitive is allowed to share a common part of its trajectory with other primitives and to specialize later.
This paper\'s intention is to adapt prediction algorithms well known in the field of time ser... more This paper\'s intention is to adapt prediction algorithms well known in the field of time series analysis to problems being faced it the field of mobile robotics and Human-Computer-Interation (HRI). The idea is to predict movement data by understanding it as time series. The prediction takes place with o black box model, which means that no further knowledge on motion dynamics is used then the post of the trajectory itself:This means, the suggested approaches are able to adapt yo different situations. Several state-of-the-art algorithms such as Local Modeling, Cluster Weighted Modeling, Echo State Networks and Autoregressive Models are evaluated and compared. For experiments, real movement trajectories of a human are used. Since mobile robots highly depend on real-time application, computing time is also considered. Experiments show that Echo State Networks and Local Models show impressive results for long term motion prediction with a prediction horizon of up to eight seconds.
Advances in Intelligent Systems and Computing, 2014
This paper aims at an approach for labeling places within a grid cell environment. For that we pr... more This paper aims at an approach for labeling places within a grid cell environment. For that we propose a method that is based on non-negative matrix factorization (NMF) to extract environment specific features from a given occupancy grid map. NMF also computes a description about where on the map these features need to be applied. We use this description after certain pre-processing steps as an input for generalized learning vector quantization (GLVQ) to achieve the classification or labeling of the grid cells. Our approach is evaluated on a standard data set from University of Freiburg, showing very promising results.
Der Beitrag stellt mit dem Projekt „FIPS“ das Konzept eines Web-basierten Praktikums mit verschie... more Der Beitrag stellt mit dem Projekt „FIPS“ das Konzept eines Web-basierten Praktikums mit verschiedenen Experimenten vor, die uber das Internet gesteuert, beobachtet und ausgewertet werden konnen. Dabei sollen neben den verschiedenen Realisierungsmoglichkeiten auch die einzelnen Schritte der Versuchsdurchfuhrung — angefangen vom Entwurf des Steuerungsalgorithmus bis hin zur web-gesteuerten Erprobung am realen Modell — anhand des Praktikumsversuches „2D-Laufkatzensteuerung“ vorgestellt werden. Darauf aufbauend werden die im Rahmen des Projektes „FIPS“ entwickelte Systemarchitektur vorgestellt sowie verschiedene Realisierungsmoglichkeiten bezuglich der Abarbeitung des entworfenen Steuerungsalgorithmus aufgezeigt und miteinander verglichen.
ABSTRACT We propose a system for enhancing a human-robot interaction system. The goal is to embra... more ABSTRACT We propose a system for enhancing a human-robot interaction system. The goal is to embrace surrounding persons into a conversation with the robot by means of establishing or keeping eye contact. For this, a simple and hence computationally efficient people tracking algorithm has been developed, which in itself is an enhancement of existing approaches. As the foundation of our approach we use a vanilla Kalman filter that explicitly models the velocity in the state space, while the observations are transformed into a unified global coordinate system. We evaluate different integration strategies for multiple sensor cues. Furthermore, we extend an algorithm for group detection to be able to recognize individuals.
2009 IEEE International Conference on Robotics and Automation, 2009
Recent advances in the field of humanoid robotics increase the complexity of the tasks that such ... more Recent advances in the field of humanoid robotics increase the complexity of the tasks that such robots can perform. This makes it increasingly difficult and inconvenient to program these tasks manually. Furthermore, humanoid robots, in contrast to industrial robots, should in the distant future behave within a social environment. Therefore, it must be possible to extend the robot's abilities in an easy and natural way. To address these requirements, this work investigates the topic of imitation learning of motor skills. The focus lies on providing a humanoid robot with the ability to learn new bimanual tasks through the observation of object trajectories. For this, an imitation learning framework is presented, which allows the robot to learn the important elements of an observed movement task by application of probabilistic encoding with Gaussian Mixture Models. The learned information is used to initialize an attractor-based movement generation algorithm that optimizes the reproduced movement towards the fulfillment of additional criteria, such as collision avoidance. Experiments performed with the humanoid robot ASIMO show that the proposed system is suitable for transferring information from a human demonstrator to the robot. These results provide a good starting point for more complex and interactive learning tasks.
ABSTRACT While he idea of tour guide robots has been addressed several times, applying a video-pr... more ABSTRACT While he idea of tour guide robots has been addressed several times, applying a video-projector based augmented reality component to this scenario is rather new. We show requirements of the localization system of the robot and how they can be fulfilled, as well as a basic system for projection correction and its suitability for this scenario.
Uploads
Papers by Sven Hellbach