Refine
Keywords
- AdaBoost (1)
- Bag of Features (1)
- Error-correcting output codes (1)
- Face normalization (1)
- Facial expression recognition (1)
- Gabor features (1)
- Human-Robot Interaction (1)
- One-vs-One multi-classification (1)
- People Detection (1)
- RGB-D (1)
This work presents a person independent pointing gesture recognition application. It uses simple but effective features for the robust tracking of the head and the hand of the user in an undefined environment. The application is able to detect if the tracking is lost and can be reinitialized automatically. The pointing gesture recognition accuracy is improved by the proposed fingertip detection algorithm and by the detection of the width of the face. The experimental evaluation with eight different subjects shows that the overall average pointing gesture recognition rate of the system for distances up to 250 cm (head to pointing target) is 86.63% (with a distance between objects of 23 cm). Considering just frontal pointing gestures for distances up to 250 cm the gesture recognition rate is 90.97% and for distances up to 194 cm even 95.31%. The average error angle is 7.28◦.
This paper presents the b-it-bots@Home team and its mobile service robot called Jenny – a service robot based on the Care-O-bot 3 platform manufactured by the Fraunhofer Institute for Manufacturing Engineering and Automation. In this paper, an overview of the robot control architecture and its capabilities is presented. The capabilities refers to the added functionalities from research and projects carried out within the Bonn-Rhein-Sieg University of Applied Science.
We present a system to automatically recognize facial expressions from static images. Our approach consists of extracting particular Gabor features from normalized face images and mapping them into three of the six basic emotions: joy, surprise and sadness, plus neutrality. Selection of the Gabor features is performed via the AdaBoost algorithm. We evaluated two learning machines (AdaBoost and Support Vector Machines), two multi-classification strategies (Error-Correcting Output Codes and One-vs-One) and two face image sizes (48 x 48 and 96 x 96). Images of the Cohn-Kanade AU-Coded Facial Expression Database were used as test bed for our research. Best results (87.14% recognition rate) were obtained using Support Vector Machines in combination with Error-Correcting Output Codes and normalized face images of 96 x 96.
The ability to detect people in domestic and unconstrained environments is crucial for every service robot. The knowledge where people are is required to perform several tasks such as navigation with dynamic obstacle avoidance and human-robot-interaction. In this paper we propose a people detection approach based on 3d data provided by a RGB-D camera. We introduce a novel 3d feature descriptor based on Local Surface Normals (LSN) which is used to learn a classifier in a supervised machine learning manner. In order to increase the systems flexibility and to detect people even under partial occlusion we introduce a top-down/bottom-up segmentation. We deployed the people detection system on a real-world service robot operating at a reasonable frame rate of 5Hz. The experimental results show that our approach is able to detect persons in various poses and motions such as sitting, walking, and running.
After introducing the waveform relaxation (WR) concept, the authors briefly describe how WR was implemented in the simulator SISAL. They analyze a refinement of WR called waveform relaxation Newton (WRN). They reveal the interrelation of both methods and introduce an improved implementation technique of the WRN algorithm. Numerical results are reported.
An efficient object perception is a crucial component of a mobile service robot. In this work we present a solution for visual categorization of objects. We developed a prototypic categorization system which classifies unknown objects based on their visual properties to a corresponding category of predefined domestic object categories. The system uses the Bag of Features approach which does not rely on global geometric object information. A major contribution of our work is the enhancement of the categorization accuracy and robustness through a selected combination of a set of supervised machine learners which are trained with visual information from object instances. Experimental results are provided which benchmark the behavior and verify the performance regarding the accuracy and robustness of the proposed system. The system is integrated on a mobile service robot to enhance its perceptual capabilities, hence computational cost and robot dependent properties are considered as essential design criteria.
Field-programmable gate arrays (FPGAs) are integrated circuits that can be designed or configured after manufacturing. They are used in many disciplines to create prototypes of hardware or in applications where hardware functionality needs to be changed more frequently. Design of new FPGA architectures requires tools that allow developers to create new structures and test those structures in order to compare the results to already established solutions. Boolean circuits, implemented on the FPGAs are compiled using hardware description languages such as Verilog or VHDL. The VTR (Verilog to Routing) CAD (Computer Aided Design) tool, compiles Verilog source code that targets specific hardware resources as FPGAs and ASICs (Application Specific Integrated Circuits). The VTR CAD tool consists of three tools: Odin II, for elaboration from Verilog to a netlist, ABC, for logic synthesis, and VPR, for physical synthesis and analysis. Odin II currently supports only a sub-set of constructs in Verilog language. This paper describes improved and expanded language support for Verilog elaboration introduced in Odin II, in order to provide developers with a tool set to assist in modern FPGA research. With this enhanced language support, a subsequent evaluation of the performance characteristics of VTR flow with a set of benchmarks that are supported by VTR is performed.
