An efficient network architecture with guaranteed QoS and very high availability for real-time on-line learning

2005 ◽  
Author(s):  
M. Gainage ◽  
M. Hayasaka ◽  
T. Miki
Keyword(s):  
Real Time ◽  
Network Architecture ◽  
High Availability ◽  
On Line ◽  
On Line Learning ◽  
Very High

MODELING AND AUGMENTING GAME ENTERTAINMENT THROUGH CHALLENGE AND CURIOSITY

2007 ◽  
Vol 16 (06) ◽  
pp. 981-999 ◽  
Author(s):  
GEORGIOS N. YANNAKAKIS ◽  
JOHN HALLAM
Keyword(s):  
Real Time ◽  
Computer Games ◽  
Mixed Reality ◽  
Computer Game ◽  
Feedforward Neural Networks ◽  
Interactive Systems ◽  
Quantitative Measurements ◽  
Player Satisfaction ◽  
On Line ◽  
On Line Learning

This paper presents quantitative measurements/metrics of qualitative entertainment features within computer game environments and proposes artificial intelligence (AI) techniques for optimizing entertainment in such interactive systems. A human-verified metric of interest (i.e. player entertainment in real-time) for predator/prey games and a neuro-evolution on-line learning (i.e. during play) approach have already been reported in the literature to serve this purpose. In this paper, an alternative quantitative approach to entertainment modeling based on psychological studies in the field of computer games is introduced and a comparative study of the two approaches is presented. Feedforward neural networks (NNs) and fuzzy-NNs are used to model player satisfaction (interest) in real-time and investigate quantitatively how the qualitative factors of challenge and curiosity contribute to human entertainment. We demonstrate that appropriate non-extreme levels of challenge and curiosity generate high values of entertainment and we project the extensibility of the approach to other genres of digital entertainment (e.g. mixed-reality interactive playgrounds).

Continuous, on-line, real-time spectral analysis of SAP signals during cardiopulmonary bypass

1995 ◽  
Vol 268 (6) ◽  
pp. H2329-H2335
Author(s):  
M. W. Yang ◽  
T. B. Kuo ◽  
S. M. Lin ◽  
K. H. Chan ◽  
S. H. Chan
Keyword(s):  
Spectral Analysis ◽  
Cardiopulmonary Bypass ◽  
Blood Vessels ◽  
Real Time ◽  
Low Frequency ◽  
Frequency Component ◽  
Very Low Frequency ◽  
Frequency Components ◽  
On Line ◽  
Very High

We communicated the application of continuous, on-line, real-time power spectral analysis of systemic arterial pressure (SAP) signals during cardiopulmonary bypass when the heart was functionally but reversibly disconnected from the blood vessels. Based on observations from 15 cases of successfully completed coronary artery bypass grafting procedures, we found that the very low (0.00-0.08 Hz), low (0.08-0.15 Hz)-, high (0.15-0.25 Hz)-, and very high (0.80-1.60 Hz) frequency components of SAP signals exhibited differential changes before, during, and after cardiopulmonary bypass. In particular, the very low-frequency component, which purportedly represents the contribution of vasomotor activity to SAP, presented only a mild decrease in power during hypothermic cardioplegia. Interestingly, the total peripheral resistance also manifested only a slight reduction during the same period. On the other hand, the low-, high-, and very high frequency components were essentially eliminated. These results unveiled an active role for the blood vessels in the maintenance of SAP during cardiopulmonary bypass, possibly as a result of a maintained vasomotor tone as reflected by the sustained very low frequency component of the SAP signals.

A Fixed Size Storage O(n3) Time Complexity Learning Algorithm for Fully Recurrent Continually Running Networks

1992 ◽  
Vol 4 (2) ◽  
pp. 243-248 ◽  
Author(s):  
Jürgen Schmidhuber
Keyword(s):  
Real Time ◽  
Time Complexity ◽  
Learning Algorithm ◽  
Fixed Size ◽  
Time Step ◽  
The Real ◽  
On Line ◽  
On Line Learning

The real-time recurrent learning (RTRL) algorithm (Robinson and Fallside 1987; Williams and Zipser 1989) requires O(n4) computations per time step, where n is the number of noninput units. I describe a method suited for on-line learning that computes exactly the same gradient and requires fixed-size storage of the same order but has an average time complexity per time step of O(n3).

scholarly journals DSNet: an efficient CNN for road scene segmentation

2020 ◽  
Vol 9 ◽  
Author(s):  
Ping-Rong Chen ◽  
Hsueh-Ming Hang ◽  
Sheng-Wei Chan ◽  
Jing-Jhih Lin
Keyword(s):  
Real Time ◽  
Network Architecture ◽  
Autonomous Driving ◽  
High Accuracy ◽  
Neural Net ◽  
Scene Segmentation ◽  
Convolutional Network ◽  
Advantages And Disadvantages ◽  
Segmentation Accuracy ◽  
Very High

Road scene understanding is a critical component in an autonomous driving system. Although the deep learning-based road scene segmentation can achieve very high accuracy, its complexity is also very high for developing real-time applications. It is challenging to design a neural net with high accuracy and low computational complexity. To address this issue, we investigate the advantages and disadvantages of several popular convolutional neural network (CNN) architectures in terms of speed, storage, and segmentation accuracy. We start from the fully convolutional network with VGG, and then we study ResNet and DenseNet. Through detailed experiments, we pick up the favorable components from the existing architectures and at the end, we construct a light-weight network architecture based on the DenseNet. Our proposed network, called DSNet, demonstrates a real-time testing (inferencing) ability (on the popular GPU platform) and it maintains an accuracy comparable with most previous systems. We test our system on several datasets including the challenging Cityscapes dataset (resolution of 1024 × 512) with an Mean Intersection over Union (mIoU) of about 69.1% and runtime of 0.0147 s/image on a single GTX 1080Ti. We also design a more accurate model but at the price of a slower speed, which has an mIoU of about 72.6% on the CamVid dataset.

THE CLUSNET ALGORITHM AND TIME SERIES PREDICTION

1993 ◽  
Vol 04 (03) ◽  
pp. 247-255 ◽  
Author(s):  
W. HSU ◽  
L. S. HSU ◽  
M. F. TENORIO
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Network Architecture ◽  
Time Series Prediction ◽  
Clustering Algorithms ◽  
Linear Method ◽  
Trade Offs ◽  
Local Linear Method ◽  
On Line ◽  
On Line Learning

This paper describes a novel neural network architecture named ClusNet. This network is designed to study the trade-offs between the simplicity of instance-based methods and the accuracy of the more computational intensive learning methods. The features that make this network different from existing learning algorithms are outlined. A simple proof of convergence of the ClusNet algorithm is given. Experimental results showing the convergence of the algorithm on a specific problem is also presented. In this paper, ClusNet is applied to predict the temporal continuation of the Mackey–Glass chaotic time series. A comparison between the results obtained with ClusNet and other neural network algorithms is made. For example, ClusNet requires one-tenth the computing resources of the instance-based local linear method for this application while achieving comparable accuracy in this task. The sensitivity of ClusNet prediction accuracies on specific clustering algorithms is examined for an application. The simplicity and fast convergence of ClusNet makes it ideal as a rapid prototyping tool for applications where on-line learning is required.

Sign in / Sign up

Export Citation Format

Share Document