A learning environment for maintenance of power equipment using Virtual Reality technology is described. the design concept is to support developing operator's mental model based on the investigations of the maint...
详细信息
A learning environment for maintenance of power equipment using Virtual Reality technology is described. the design concept is to support developing operator's mental model based on the investigations of the maintenance expertise and insights on human cognition. Currently, the system which is considered to be effective through demonstration as test usage is being evaluated.
In this paper an extension to the standard error back-propagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to trained for context dependent information. the context dep...
详细信息
In this paper an extension to the standard error back-propagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to trained for context dependent information. the context dependent learning is realised by using a different error function (called Average Risk: AVR) in stead of the sum of squared errors (SQE) normally used in error backpropagation and by adapting the update rules. It is shown that for applications where this context dependent information is important, a major improvement in performance is obtained.
In this paper a technique for eradicating the problem of lip synchronization has been discussed. When motion in an image sequence is large, the present day videophone systems are unable to maintain synchronization of ...
详细信息
In this paper a technique for eradicating the problem of lip synchronization has been discussed. When motion in an image sequence is large, the present day videophone systems are unable to maintain synchronization of the lips and the corresponding acoustic signal. By exploiting the correlation found between the speech and image data, the problem of lip synchronization is addressed and the mouth and speech signals may be transmitted in the same packet.
this paper concentrates on the estimation of dense depth maps from sequences of frames acquired by a sensor in controlled motion. the work addresses the computation of depth estimates at each time instant as well as t...
详细信息
this paper concentrates on the estimation of dense depth maps from sequences of frames acquired by a sensor in controlled motion. the work addresses the computation of depth estimates at each time instant as well as the dynamic fusion of estimates in time.
the objective of this study is to find an appropriate method of improving spectral image classifications. To fulfill this objective, a slightly modified k-Nearest-Neighbour strategy is proposed for the calculation of ...
详细信息
the objective of this study is to find an appropriate method of improving spectral image classifications. To fulfill this objective, a slightly modified k-Nearest-Neighbour strategy is proposed for the calculation of feature-probability densities. In addition, the method of using spatially distributed prior probabilities is reviewed and shown how it can be perfectly combined withthe proposed method. By conducting several experiments, it is found that the proposed strategy serves the purpose of accurately estimating class-areas, as well as increasing classification accuracies to the same extent as the 'pixel prior' method does.
A range image consists of noisy, discrete samples of object surfaces. Perception of surfaces plays a key role in range image understanding and 3D object recognition. In this paper a statistically robust and computatio...
详细信息
A range image consists of noisy, discrete samples of object surfaces. Perception of surfaces plays a key role in range image understanding and 3D object recognition. In this paper a statistically robust and computationally efficient function approximation scheme for surface segmentation is developed. In the existing surface characterization techniques, surface completion is done after segmentation, using heuristics and rules to piece together parts of the complete surface. Our novel method generates the complete surface hypotheses in parameter space in one step. Object segmentation is automatically achieved in the reduced parameter space instead of the image space. the algorithms are tested and supported by extensive experiments on real and synthetic depth maps that exhibit surface coherence property.
Practical tools for knowledge discovery from databases must be efficient enough to handle large data sets found in commercial environments. Attribute-oriented induction has proved to be a useful method for knowledge d...
详细信息
Practical tools for knowledge discovery from databases must be efficient enough to handle large data sets found in commercial environments. Attribute-oriented induction has proved to be a useful method for knowledge discovery. three algorithms are AOI, LCHR and GDBR. We have implemented efficient versions of each algorithm and empirically compared them on large commercial data sets. these tests show that GDBR is consistently faster than AOI and LCHR. GDBR's times increase linearly with increased input size, while times for AOI and LCHR increase non-linearly when memory is exceeded. through better memory management, however, AOI can be improved to provide some advantages.
Visual occlusion events constitute a major source of depth information. this paper presents a self-organizing neural network that learns to detect, represent, and predict the visibility and invisibility relationships ...
详细信息
When the dimension N of the input vector is much larger than the number M of different training patterns to be learned, a one-layered, hard-limited perceptron with N input nodes and P neurons (P > equals Log2M)...
详细信息
ISBN:
(纸本)0819415472
When the dimension N of the input vector is much larger than the number M of different training patterns to be learned, a one-layered, hard-limited perceptron with N input nodes and P neurons (P > equals Log2M) is generally sufficient to accomplish the learning- recognition task. the recognition should be very robust and very fast if an optimum noniterative learning scheme is applied to the perceptron learning process. this paper concentrates at the discussion of two special characteristics of this novel patternrecognition system: the automatic feature extraction and the automatic feature competition. An unedited video movie recorded on a series of learning-recognition experiments may demonstrate these characteristics of the novel system in real time.
An attempt is made to find a new way for better diagnosis of hepatisis through application of artificial neural network theory. learning from a given sample set, the neural network is used to establish a nonlinear map...
详细信息
ISBN:
(纸本)0819415472
An attempt is made to find a new way for better diagnosis of hepatisis through application of artificial neural network theory. learning from a given sample set, the neural network is used to establish a nonlinear mapping between various factors, such as symptoms, signs, and laboratorial experiments, and diagnosis of hepatisis. It is proved that the used network and values of weight after learning are available to the identification of equivalent class of a new pattern of hepatisis. In this paper, the knowledge learning and learning algorithms used in diagnosis are mainly discussed, an optimal generalization algorithm based on the error decrease algorithm and used to train multilayer feedforward is presented; meanwhile, the application results and their effectiveness are introduced.
暂无评论