In Automatic Speech Recognition (ASR) the non-linear data projection provided by a one hidden layer Multilayer Perceptron (MLP), trained to recognize phonemes, and has previous experiments to provide feature enhancement substantially increased ASR performance, especially in noise. Previous attempts to apply an analogous approach to speaker identification have not succeeded in improving performance, except by combining MLP processed features with other features. We present test results for the TIMIT database which show that the advantage of MLP preprocessing for open set speaker identification increases with the number of speakers used to train the MLP and that improved identification is obtained as this number increases beyond sixty.

Digital image is widely used in computer applications. This paper introduces a proposed method of image zooming based upon inverse slantlet transform and image scaling. Slantlet transform (SLT) is based on the principle of designing different filters for different scales.

      First we apply SLT on color image, the idea of transform color image into slant, where large coefficients are mainly the   signal and smaller one represent the noise. By suitably modifying these coefficients , using scaling up image by  box and Bartlett filters so that the image scales up to 2X2 and then inverse slantlet transform from modifying coefficients using to the reconstructed image .

  &nbs

This paper presents a hierarchical two-stage outdoor scene classification method using multi-classes of Support Vector Machine (SVM). In this proposed method, the gist feature of all the images in the database is extracted first to obtain the feature vectors. The image of database is classified into eight outdoor scenes classes, four manmade scenes and four natural scenes. Second, a hierarchical classification is applied, where the first stage classifies all manmade scene classes against all natural scene classes, while the second stage of a hierarchical classification classifies the outputs of first stage into either one of the four manmade scene classes or natural scene classes. Binary SVM and multi-classes SVMs are employed in the fir

Feature extraction provide a quick process for extracting object from remote sensing data (images) saving time to urban planner or GIS user from digitizing hundreds of time by hand. In the present work manual, rule based, and classification methods have been applied. And using an object- based approach to classify imagery. From the result, we obtained that each method is suitable for extraction depending on the properties of the object, for example, manual method is convenient for object, which is clear, and have sufficient area, also choosing scale and merge level have significant effect on the classification process and the accuracy of object extraction. Also from the results the rule-based method is more suitable method for extracting

Median filter is adopted to match the noise statistics of the degradation seeking good quality smoothing images. Two methods are suggested in this paper(Pentagonal-Hexagonal mask and Scan Window Mask), the study involved modified median filter for improving noise suppression, the modification is considered toward more reliable results. Modification median filter (Pentagonal-Hexagonal mask) was found gave better results (qualitatively and quantitatively ) than classical median filters and another suggested method (Scan Window Mask), but this will be on the account of the time required. But sometimes when the noise is line type the cross 3x3 filter preferred to another one Pentagonal-Hexagonal with few variation. Scan Window Mask gave bett

     Improving the performance of visual computing systems is achieved by removing unwanted reflections from a picture captured in front of a glass. Reflection and transmission layers are superimposed in a linear form at the reflected photographs. Decomposing an image into these layers is often a difficult task. Plentiful classical separation methods are available in the literature which either works on a single image or requires multiple images. The major step in reflection removal is the detection of reflection and background edges. Separation of the background and reflection layers is depended on edge categorization results. In this paper a wavelet transform is used as a prior estimation of background edges to sepa

The objective of this work is to design and implement a cryptography system that enables the sender to send message through any channel (even if this channel is insecure) and the receiver to decrypt the received message without allowing any intruder to break the system and extracting the secret information. In this work, we implement an interaction between the feedforward neural network and the stream cipher, so the secret message will be encrypted by unsupervised neural network method in addition to the first encryption process which is performed by the stream cipher method. The security of any cipher system depends on the security of the related keys (that are used by the encryption and the decryption processes) and their corresponding le