In this article we present an automatic on-line color calibration system that makes extensive use... more In this article we present an automatic on-line color calibration system that makes extensive use of the spatial relationships between color classes in the color space. First, we introduce the definition of class-relative color spaces, where classes are represented in terms of their spatial relation to a base color class. Then, using class-relative color spaces, the system is able to remap classes from the already trained ones, which gives a starting point for training the remaining classes. The color-calibrating system also uses a feedback from the detected objects using the remapped (or partially trained) classes. As a result, the system is able to generate a complete color look-up table from scratch, and to adapt quickly to severe lighting condition changes. A particularity of our system is that it does not need to solve the natural ambiguity in color classes’ intersections, but it is able to keep and use it during color segmentation using the concept of soft-colors.
A mobile robot has always uncertainty about the world model. Reducing this uncertainty is very ha... more A mobile robot has always uncertainty about the world model. Reducing this uncertainty is very hard because there is a huge amount of information and the robot must focus on the most relevant one. The selection of the most relevant information must be based on the task the robot is executing, but there could be several sources of information where the robot would like to focus on. This is also true in robot soccer where the robot must pay attention to landmarks in order to self-localize and to the ball and robots in order to follow the status of the game. In the presented work, an explicitly task oriented probabilistic active vision system is proposed. The system tries to minimize the most relevant components of the uncertainty for the task that is been performed and it is explicitly task oriented in the sense that it explicitly considers a task specific value function. As a result, the system estimates the convenience of looking towards each of the available objects. As a test-bed for the presented active vision approach, we selected a robot soccer attention problem: goal covering by a goalie player.
A real-time hybrid control architecture for biped humanoid robots is proposed. The architecture i... more A real-time hybrid control architecture for biped humanoid robots is proposed. The architecture is modular and hierarchical. The main robot’s functionalities are organized in four parallel modules: perception, actuation, world-modeling, and hybrid control. Hybrid control is divided in three behavior-based hierarchical layers: the planning layer, the deliberative layer, and the reactive layer, which work in parallel and have very different response speeds and planning capabilities. The architecture allows: (1) the coordination of multiple robots and the execution of group behaviors without disturbing the robot’s reactivity and responsivity, which is very relevant for biped humanoid robots whose gait control requires real-time processing. (2) The straightforward management of the robot’s resources using resource multiplexers. (3) The integration of active vision mechanisms in the reactive layer under control of behavior-dependant value functions from the deliberative layer. This adds flexibility in the implementation of complex functionalities, such as the ones required for playing soccer in robot teams. The architecture is validated using simulated and real Nao humanoid robots. Passive and active behaviors are tested in simulated and real robot soccer setups. In addition, the ability to execute group behaviors in real- time is tested in international robot soccer competitions.
Recognition of relevant game field objects, such as the ball and landmarks, is usually based upon... more Recognition of relevant game field objects, such as the ball and landmarks, is usually based upon the application of a set of decision rules over candidate image regions. Rule selection and parameters tuning are often arbitrarily done. We propose a method for evolving the selection of these rules as well as their parameters with basis on real game field images, and a supervised learning approach. The learning approach is implemented using genetic algorithms. Results of the application of our method are presented.
Decision making is an important issue in robot soccer, which has not been investigated deeply eno... more Decision making is an important issue in robot soccer, which has not been investigated deeply enough by the RoboCup research community. This paper proposes a probabilistic approach to decision making. The proposed methodology is based on the maximization of a game situation score function, which generalizes the concept of accomplishing different game objectives as: passing, scoring a goal, clearing the ball, etc. The methodology includes a quantitative method for evaluating the game situation score. Experimental results in a high-level strategy simulator, which runs our four-legged code in simulated AIBOs’ robots, show a noticeable improvement in the scoring effectiveness achieved by a team that uses the proposed approach for making decisions.
Having as a main motivation the development of robust and high performing robot vision systems th... more Having as a main motivation the development of robust and high performing robot vision systems that can operate in dynamic environments, we propose a bayesian spatiotemporal context-based vision system for a mobile robot with a mobile camera, which uses three different context-coherence instances: current frame coherence, last frame coherence and high level tracking coherence (coherence with tracked objects). We choose as a first application for this vision system, the detection of static objects in the RoboCup Standard Platform League domain. The system has been validated using real video sequences and has presented satisfactory results. A relevant conclusion is that the last frame coherence appears to be not very important in the tested cases, while the coherence with the tracked objects appears to be the most important context level considered.
In this article we present an automatic on-line color calibration system that makes extensive use... more In this article we present an automatic on-line color calibration system that makes extensive use of the spatial relationships between color classes in the color space. First, we introduce the definition of class-relative color spaces, where classes are represented in terms of their spatial relation to a base color class. Then, using class-relative color spaces, the system is able to remap classes from the already trained ones, which gives a starting point for training the remaining classes. The color-calibrating system also uses a feedback from the detected objects using the remapped (or partially trained) classes. As a result, the system is able to generate a complete color look-up table from scratch, and to adapt quickly to severe lighting condition changes. A particularity of our system is that it does not need to solve the natural ambiguity in color classes’ intersections, but it is able to keep and use it during color segmentation using the concept of soft-colors.
A mobile robot has always uncertainty about the world model. Reducing this uncertainty is very ha... more A mobile robot has always uncertainty about the world model. Reducing this uncertainty is very hard because there is a huge amount of information and the robot must focus on the most relevant one. The selection of the most relevant information must be based on the task the robot is executing, but there could be several sources of information where the robot would like to focus on. This is also true in robot soccer where the robot must pay attention to landmarks in order to self-localize and to the ball and robots in order to follow the status of the game. In the presented work, an explicitly task oriented probabilistic active vision system is proposed. The system tries to minimize the most relevant components of the uncertainty for the task that is been performed and it is explicitly task oriented in the sense that it explicitly considers a task specific value function. As a result, the system estimates the convenience of looking towards each of the available objects. As a test-bed for the presented active vision approach, we selected a robot soccer attention problem: goal covering by a goalie player.
A real-time hybrid control architecture for biped humanoid robots is proposed. The architecture i... more A real-time hybrid control architecture for biped humanoid robots is proposed. The architecture is modular and hierarchical. The main robot’s functionalities are organized in four parallel modules: perception, actuation, world-modeling, and hybrid control. Hybrid control is divided in three behavior-based hierarchical layers: the planning layer, the deliberative layer, and the reactive layer, which work in parallel and have very different response speeds and planning capabilities. The architecture allows: (1) the coordination of multiple robots and the execution of group behaviors without disturbing the robot’s reactivity and responsivity, which is very relevant for biped humanoid robots whose gait control requires real-time processing. (2) The straightforward management of the robot’s resources using resource multiplexers. (3) The integration of active vision mechanisms in the reactive layer under control of behavior-dependant value functions from the deliberative layer. This adds flexibility in the implementation of complex functionalities, such as the ones required for playing soccer in robot teams. The architecture is validated using simulated and real Nao humanoid robots. Passive and active behaviors are tested in simulated and real robot soccer setups. In addition, the ability to execute group behaviors in real- time is tested in international robot soccer competitions.
Recognition of relevant game field objects, such as the ball and landmarks, is usually based upon... more Recognition of relevant game field objects, such as the ball and landmarks, is usually based upon the application of a set of decision rules over candidate image regions. Rule selection and parameters tuning are often arbitrarily done. We propose a method for evolving the selection of these rules as well as their parameters with basis on real game field images, and a supervised learning approach. The learning approach is implemented using genetic algorithms. Results of the application of our method are presented.
Decision making is an important issue in robot soccer, which has not been investigated deeply eno... more Decision making is an important issue in robot soccer, which has not been investigated deeply enough by the RoboCup research community. This paper proposes a probabilistic approach to decision making. The proposed methodology is based on the maximization of a game situation score function, which generalizes the concept of accomplishing different game objectives as: passing, scoring a goal, clearing the ball, etc. The methodology includes a quantitative method for evaluating the game situation score. Experimental results in a high-level strategy simulator, which runs our four-legged code in simulated AIBOs’ robots, show a noticeable improvement in the scoring effectiveness achieved by a team that uses the proposed approach for making decisions.
Having as a main motivation the development of robust and high performing robot vision systems th... more Having as a main motivation the development of robust and high performing robot vision systems that can operate in dynamic environments, we propose a bayesian spatiotemporal context-based vision system for a mobile robot with a mobile camera, which uses three different context-coherence instances: current frame coherence, last frame coherence and high level tracking coherence (coherence with tracked objects). We choose as a first application for this vision system, the detection of static objects in the RoboCup Standard Platform League domain. The system has been validated using real video sequences and has presented satisfactory results. A relevant conclusion is that the last frame coherence appears to be not very important in the tested cases, while the coherence with the tracked objects appears to be the most important context level considered.
Uploads
Papers by pablo guerrero