Speech recognition over lines forms an integral part of various applications of large vocabulary continuous speech recognition (LVCSR). This paper describes an implementation system completely in software form to prod...
详细信息
ISBN:
(纸本)078037925X
Speech recognition over lines forms an integral part of various applications of large vocabulary continuous speech recognition (LVCSR). This paper describes an implementation system completely in software form to produce simulated telephone data starting from clean databases. Filters adopted in this system are well-designed to simulate the frequency properties of analogue transmission equipments in telephone connection. A speech recognizer was, trained from speech data extracted from clean corpus piped through a hardware simulator. The recognition performances are evaluated on a real telephone speech set and several test sets simulated from the clean database for test use. The experiments verified the effectiveness and feasibility of software simulation from a recognition testing point of view, and the results showed using simulated data derived from clean corpus could achieve the same recognition performance as real telephone speech.
A scale-accurate robotic model of a bat biosonar head which reproduces all known functional characteristics of its biological paragon is being build in the CIRCE project. The heads of bats are a set of intelligent, ad...
详细信息
ISBN:
(纸本)0780375793
A scale-accurate robotic model of a bat biosonar head which reproduces all known functional characteristics of its biological paragon is being build in the CIRCE project. The heads of bats are a set of intelligent, adaptively controlled sonar antennae, where signalprocessing is carried out both in the acoustic and in the neural domain. The primary project goal is to identify hypotheses of high explanatory power for the function of these biological systems. The progress report given here outlines involved engineering challenges and explains design decisions in the areas of ultrasonic transducers, antennae shapes, actuation and neuromimetic signalprocessing.
Visual sensing for robotics has been around for decades, but our under-standing of a timing model remains crude. By timing model, we refer to the delays (processing lag and motion lag) between "reality" (whe...
详细信息
ISBN:
(纸本)0780378601
Visual sensing for robotics has been around for decades, but our under-standing of a timing model remains crude. By timing model, we refer to the delays (processing lag and motion lag) between "reality" (when a part is sensed), through data processing (the processing of image data to determine part position and orientation), through control (the computation and initiation of robot motion), through "arrival" (when the robot reaches "reality"). In this work we introduce a timing model where sensing and control operate asynchronously. We apply this model to a robotic workcell consisting of a Staubli RX-130 industrial robot manipulator, a network of six cameras for sensing, and an off-the-shelf Adept MV-19 controller. We demonstrate some experiments to show how the model can be applied.
Tradition single-model adaptive signalprocessing system faces many difficulties when it is applied in complex signals processing and different conditions, such as constant frequency wide-band signal, fast-varying sig...
详细信息
In this paper, audio servo robot system which was a servo system using audio information as feedback signal was considered. Since the sampling rate of audio signal and that of robot controller are usually different, t...
详细信息
ISBN:
(纸本)0780378601
In this paper, audio servo robot system which was a servo system using audio information as feedback signal was considered. Since the sampling rate of audio signal and that of robot controller are usually different, the system is able to be treated as a multirate system. Since it is general that the sampling rate of the output is faster than that of the input, the controller can obtain more information than a single-rate servo system. By using this property, an adaptive audio servo system was proposed. Results of a computer simulation and a experiment showed effectiveness of the method.
Research in the MIT Media Laboratory's Object-Based Media Group is looking to solve multimedia signalprocessing problems through the creation of ecosystems of intelligent devices that organize themselves to distr...
详细信息
ISBN:
(纸本)0780377249
Research in the MIT Media Laboratory's Object-Based Media Group is looking to solve multimedia signalprocessing problems through the creation of ecosystems of intelligent devices that organize themselves to distribute information and to divide large tasks. I examine the technological trends that make this approach possible and desirable, and briefly introduce three example systems.
The proceedings contains 166 papers from the 2003ieee/RSJ internationalconference on intelligent Robots and systems (IROS 2003) Volume 2. The topics discussed include: an active artificial whisker array for texture ...
详细信息
The proceedings contains 166 papers from the 2003ieee/RSJ internationalconference on intelligent Robots and systems (IROS 2003) Volume 2. The topics discussed include: an active artificial whisker array for texture discrimination;a signal separation technique which can be applied to the ears of robots;task driven dynamic QoS based bandwidth allocation for real-time teleoperation via the internet;multi-objective optimal robot path planning in manufacturing;approaches for heuristically biasing RRT growth;active learning of a mobile robot based on percieving-acting cycle;and spiking neural building block robot with Hebbian learning.
In the conventional non-contact type extensometer, the deformation of a material is obtained on the time basis which requires high accurate scanning linear velocity. In this paper, a new method of the optical laser ex...
详细信息
ISBN:
(纸本)0780377591
In the conventional non-contact type extensometer, the deformation of a material is obtained on the time basis which requires high accurate scanning linear velocity. In this paper, a new method of the optical laser extensometer is proposed by using a position sensitive detector (PSD) which gives a direct information of the linear displacement of deformation of a material. The distance between the reflection tapes on the specimen can be obtained by using the signals from the photo-TR and the PSD while the laser beam scans along the specimen. The optical system of the proposed system, signalprocessing, control of the reflection motor, and experimental result are presented.
Legged walking robots are able to not only move on irregular terrain but also change its posture. or example, the robots can pass under above head obstacles with crouching. These advantages are particularly effective ...
详细信息
ISBN:
(纸本)078037925X
Legged walking robots are able to not only move on irregular terrain but also change its posture. or example, the robots can pass under above head obstacles with crouching. These advantages are particularly effective in three dimensional complex environment which wheeled robots cannot move on. However, some of the issues of a walking robot include instability and workspace limitation. All of these impact on the realization of applications. We have studied for the improvement of them and deadlock-less gait pattern, called Trajectory Following Gait (TFG), has been proposed. Here, the advantages of walking robots are remarked again. In order to get the most out of these advantages, an environment recognition is absolutely imperative. Applying the vision sensor to the walking robot, however, is difficult because misconception can be occurred easily by the rocking vibration. Although rough surrounding information for global planning should be recognized by the visual sensor, more reliable sensing is desirable by the local one, i.e. steps and above head obstacles. This paper presents the reliable sensing method for a local obstacle recognition using distance sensors. The overcoming locomotion and a posture control method are also proposed. Finally, the local recognition availability is verified by some experiments.
The robotic architecture framework standards, as required by the US Department of Defense (DoD), were discussed. The defense domain requirements were based on obtaining the architectures that enable the collection, pr...
详细信息
The robotic architecture framework standards, as required by the US Department of Defense (DoD), were discussed. The defense domain requirements were based on obtaining the architectures that enable the collection, processing, analysis, and integration of information and technology (IT) features. The dissemination and reusability features of the the framework provide aid in achieving interoperability objectives. In this regard, the architectural framework based on Command, Control, Communications, Computers, Intelligence, Surveillance and Reconnaissance (C4ISR) system modeling was discussed.
暂无评论