Showing papers by "Lakhmi C. Jain published in 2002"

Self-Organizing Neural Networks

Abstract: Learning algorithms which were considered for a single perceptron, linear adaline, and multilayer perceptron belong to the class of supervised learning algorithms. In this case the training data is divided into input signals, x(n), and target signals, d(n). A typical learning algorithm is driven by error signals ε(n) which are the differences between the actual network output, y(n), and the desire (or target) output for a given input. For a pattern learning, we can express the weight update in the following general form ∆w(n) = L(w(n), x(n), ε(n)) where L represents a learning algorithm. If we say that a neural network can describe a model of data, then a multilayer perceptron describes the data in a form of a hypersurface which approximates a functional relationship between x(n), and d(n).

Adaptive Database Learning in Decision Support Systems Using Evolutionary Fuzzy Systems: A Generic Framework

Abstract: Normally a decision support system is build to solve problem where multi-criteria decisions are involved. The database is the vital part of the decision support containing the information or data that is used in decision-making process. This is the field where engineers and scientists try to apply several heuristics and soft computing techniques such as learning, search and modelling the imprecise information to obtain optimal decisions. In this paper, we present the proposed a generic framework using evolutionary fuzzy systems to obtain decision rules automatically and the usage of fuzzy inference system to process the imprecise information. Some simulation results demonstrating the difficulties to decide the optimal quantity of membership functions, shape and parameters are also provided.

Adaptation of a Mamdani Fuzzy Inference System Using Neuro-Genetic Approach for Tactical Air Combat Decision Support System

Abstract: Normally a decision support system is build to solve problems where multi-criteria decisions are involved. The knowledge base is the vital part of the decision support system containing the information or data that is used in decision-making process. This is the field where engineers and scientists have applied several intelligent techniques and heuristics to obtain optimal decisions from imprecise information. In this paper, we present a hybrid neurogenetic learning approach for the adaptation of a Mamdani fuzzy inference system for the Tactical Air Combat Decision Support System (TACDSS). Some simulation results demonstrating the different learning techniques are also provided.

Teaming Human and Machine: A Conceptual Framework

Abstract: Advances in automation and especially artificial intelligence have enabled the formation of rather unique teams with human and (electronic) machine members. This paper proposes a conceptual framework for teaming human and machine. The basis of this framework will be the introduction of the machine into the traditional situation where the human is solely responsible for managing, control and execution of all activities. Focus will be on the identification and classification of activities to be allocated to the machine. Task management and coordination between human and machine will be identified as a specific area of research and design concern.

A multi-user based watermarking system with two-security-level keys

Abstract: This paper presents a new watermarking scheme to reduce the detection error probability through removal of local mean values of an original signal. It will be analytically proven that the removal process plays role in reducing the interference between an original signal and a watermark to be embedded. This is simply based on the orthogonality of the DC signal and the AC signal. As a result, the process improves the ability of the right detection. The proposed method is analytically and empirically evaluated with no attack as well as JPEG compression attacks.

Neural Techniques in Logo Recognition

Abstract: A novel scheme is presented to detect and recognise a logo in a given document(s). Another area of interest will be dealing with distorted logos. This refers to logos, which are scaled, rotated, and have a brightness or contrast variation from the original logo. The system recognises these logos and makes correct judgements regarding their identity. The success rate for this system is about 75 to 80 percent

Fusion of Fuzzy System and Conventional Technique to Evaluate Weather and Terrain Effects on the Vehicle Operations

Abstract: This research paper presents the implementation of fuzzy logic system tools with the combination of conventional technique in evaluating the best course of action for directing military vehicles from one designated area to the final destination in quickest way. The difficulty levels between the road network nodes are weather and terrain factors, which are presented on the GIS maps. The accurate performance of the combined technique in decision making will depend on the expert’s knowledge and information supplied at the time of evaluating. The GIS model, fuzzy logic system, parameters considered for weather and terrain condition, vehicle selection, and conventional methodology will be discussed. The area between start and finish will be assumed in the 25 square Kms map area

A multistage VQ based watermarking technique with fake watermarks

Abstract: Digital watermarking is a technology proposed to help address the concern of copyright protection for digital content. To facilitate tracing of copyright violators, different watermarks carrying information about the transaction or content recipient can be embedded into multimedia content before distribution. Such form of "personalised" watermark is called "fingerprint". A powerful attack against digital fingerprinting is the collusion attack, in which different fingerprinted copies of same host data are jointly processed to remove the fingerprints or hinder their detection. This paper first studies a number of existing collusion attack schemes against image fingerprinting. A new collusion attack scheme is then proposed and evaluated, both analytically and empirically. Attack performance in terms of fingerprint detectability and visual quality degradation after attack is assessed. Results obtained from spread spectrum fingerprinting experiments show that the proposed attack can impede fingerprint detection using as few as three fingerprinted images without introducing noticeable visual degradation, hence it is more powerful than those reported in literature. It is also found that increasing the fingerprint embedding strength and spreading factor do not help resist such malicious attacks.

An Integration of Fuzzy and Two-Valued Logics on Natural Language Semantics

Abstract: This paper presents a proposal of an integration of fuzzy and two-valued logic, based on natural language semantics. We introduce a simple semantics of natural language which can treat fuzzy and two-valued logics, both of which are essential in natural language sentences. As an important example, this paper focuses on spatial relations, or scene generation system which synthesizes images of the meaning derived from natural language semantics. In these cases, as discussed below, it is better not to treat two-valued logic as a special case of fuzzy logic.

Virtual Al Classroom: A Proposal

Abstract: The combination of computers and electronic communication has the power to dramatically enhance the productivity of researchers/educators in a given area. A major step towards realizing that potential comes from combining the interests of scientific community to create an integrated common resource and a communication system to support scientific collaboration. This paper proposes a common web-based AI resource that is designed as a virtual class, and computer-supported collaborative work environments.

Perceptual Grouping of Contours via Gated Diffusion of Boundary Signals

Abstract: In this paper we present a vision based neural network architecture that models perceptual grouping of boundaries The network employs a novel concept of contour interaction via diffusion of boundary signals from edge pre-processed images This diffusion process is gated by a feedback signal from a layer of neurons, which are sensitive to the direction of surface The network is based on human pre-attentive vision and in particular on the recent neuro-physiological findings that cells in extrastriate cortex V2 perform the coding of border ownership The proposed mechanism thus leads to grouping of boundaries with surfaces that own them The model is implemented for figure-ground separation and in this paper we demonstrate its performance of perceptual grouping by testing it on a number of synthetic images