In this work, a neural structure has been implemented into a device based on the new tendencies on hardware integration, for motor-control in multisensorial anthropomorphic robotic systems. This implementation gives a...
详细信息
In this work, a neural structure has been implemented into a device based on the new tendencies on hardware integration, for motor-control in multisensorial anthropomorphic robotic systems. This implementation gives a solution to the problem of physic integration of biologically inspired control hierarchies in a robotic headarm installation for robotic reaching tasks. The complete architecture has been implemented on a electronic board connected to a PC computer through a PCI interface. The hardware structure consists on two block, one for the working phase of the system, and other for the learning and supervision phase of the system. These two blocks have been implemented with different technologies based on DSP processors and FPGAs. The algorithms implemented on DSPs have the function of updating the neural network on the FPGA, supervise the well-working of the algorithm implemented on FPGA and introduce corrections when the neural network produces results with little errors. The neural network has been implemented on FPGA and implements the spatial-motor transformations of the robotic platform. It is programmed and updated by the supervisor implemented on the DSP processor.
In autonomous robotics, so-called artificial potential fields are often used to plan and control the motion of a physical robot. In this paper, we propose to use an artificial electric field to address the problem or ...
详细信息
In this paper, we compare three representative modeling approaches, namely the multiple-prototype-based template matching approach, the subspace approach and the continuous density hidden Markov model approach for lar...
详细信息
Embedded systems are typically designed for one or a few target applications, allowing for customization of the system architecture for the desired system goals such as performance, power and cost. The memory subsyste...
详细信息
We apply fillets-smoothing of sharp angles at the joints-between the connections and nodes of graph-like diagrams. In situations where the graph layout is constrained, e.g. Focus+Context views or views where the coord...
详细信息
There is growing interest in accessing, relating, and combining data from multiple sources on the Web. Today, web users access the web through two dominant interfaces: clicking on hyperlinks (browsing) and searching v...
详细信息
ISBN:
(纸本)1853129070
There is growing interest in accessing, relating, and combining data from multiple sources on the Web. Today, web users access the web through two dominant interfaces: clicking on hyperlinks (browsing) and searching via keyword queries (crawling). This process is often time-consuming and unsatisfactory because of the many inaccurate and irrelevant results that are returned. Better support is needed for expressing one's information requirements and dealing in more structured and systematic ways with one's search results. This paper presents ALII, our proposal for providing this support. ALII is an Active Logic based framework for Information integration, which can utilize a compact context representation and constructs a hierarchy model of query and web pages. The ALII project aims developing methods and tools for a model based, semantic integration of information sources on web pages.
This paper proposes a nonlinear oscillator network LSI for image region extraction as a preprocessing of intelligent image recognition. The processing can be achieved by using the nonlinear characteristics of a nonlin...
详细信息
In our earlier work, we developed the Broadcast Incremental Power (BIP) algorithm, which is a centralized heuristic for energy efficient broadcasting of source-initiated session-based traffic in wireless networks. Thi...
详细信息
In our earlier work, we developed the Broadcast Incremental Power (BIP) algorithm, which is a centralized heuristic for energy efficient broadcasting of source-initiated session-based traffic in wireless networks. This algorithm, which exploits the characteristics of the wireless channel, was shown to perform better than adaptations of conventional algorithms that were originally developed for wired networks. However, as a consequence of its centralized nature, it is "expensive" in terms of both communication and computation requirements. In this paper, we develop two distributed versions of BIP, and compare their performance to that of centralized BIP and to an algorithm based on the minimum-cost spanning tree formulation.
As technology changes, so does the area of humancomputer interaction. HCI education must continuously change to meet the new challenges to user interaction. The World Wide Web and other distributed networks, hand-held...
详细信息
ISBN:
(纸本)1581134541
As technology changes, so does the area of humancomputer interaction. HCI education must continuously change to meet the new challenges to user interaction. The World Wide Web and other distributed networks, hand-held devices, and embedded computing all present new challenges for user-centered design methods, usability testing, and other forms of evaluation. In addition, as more people use technology, the diversity of users increases, requiring increased attention to concepts such as accessibility and universal usability. This panel will address the challenges of keeping HCI education up-to-date and offer approaches that have been successfully used. The four major topics addressed by the panel will be 1) the challenge of rapidly changing technology, 2) new methods for usercentered design, 3) student involvement with users, and 4) balancing HCI theory and HCI practice.
An intelligent controller for the trajectorycontrol of the free disc (the simplified model of thesatellite m 2-dimension space) is proposed in this *** a classifier is constructed relying on the dynamicclustering algo...
详细信息
An intelligent controller for the trajectorycontrol of the free disc (the simplified model of thesatellite m 2-dimension space) is proposed in this *** a classifier is constructed relying on the dynamicclustering algorithm. Then, a back propagation neuralnetwork is trained to learn the inverse dynamics modelof the free disc. The newly proposed method has theadvantage of simplicity and intelligence for thetrajectory control of the free disc.
暂无评论