The automatic estimation of speaker characteristics, such as height, age, and gender, has various applications in forensics, surveillance, customer service, and many human-robot interaction applications. These applications are often required to produce a response promptly. This work proposes a novel approach to speaker profiling by combining filter bank initializations, such as continuous wavelets and gammatone filter banks, with one-dimensional (1D) convolutional neural networks (CNN) and residual blocks. The proposed end-to-end model goes from the raw waveform to an estimated height, age, and gender of the speaker by learning speaker representation directly from the audio signal without relying on handcrafted and pre-computed acou

New speaker identification test’s feature, extracted from the differentiated form of the wave file, is presented. Differentiation operation is performed by an operator similar to the Laplacian operator. From the differentiated record’s, two parametric measures have been extracted and used as identifiers for the speaker; i.e. mean-value and number of zero-crossing points.

Offline handwritten signature is a type of behavioral biometric-based on an image. Its problem is the accuracy of the verification because once an individual signs, he/she seldom signs the same signature. This is referred to as intra-user variability. This research aims to improve the recognition accuracy of the offline signature. The proposed method is presented by using both signature length normalization and histogram orientation gradient (HOG) for the reason of accuracy improving. In terms of verification, a deep-learning technique using a convolution neural network (CNN) is exploited for building the reference model for a future prediction. Experiments are conducted by utilizing 4,000 genuine as well as 2,000 skilled forged signatu

Methods of speech recognition have been the subject of several studies over the past decade. Speech recognition has been one of the most exciting areas of the signal processing. Mixed transform is a useful tool for speech signal processing; it is developed for its abilities of improvement in feature extraction. Speech recognition includes three important stages, preprocessing, feature extraction, and classification. Recognition accuracy is so affected by the features extraction stage; therefore different models of mixed transform for feature extraction were proposed. The properties of the recorded isolated word will be 1-D, which achieve the conversion of each 1-D word into a 2-D form. The second step of the word recognizer requires, the

Recognizing speech emotions is an important subject in pattern recognition. This work is about studying the effect of extracting the minimum possible number of features on the speech emotion recognition (SER) system. In this paper, three experiments performed to reach the best way that gives good accuracy. The first one extracting only three features: zero crossing rate (ZCR), mean, and standard deviation (SD) from emotional speech samples, the second one extracting only the first 12 Mel frequency cepstral coefficient (MFCC) features, and the last experiment applying feature fusion between the mentioned features. In all experiments, the features are classified using five types of classification techniques, which are the Random Forest (RF),

As s widely use of exchanging private information in various communication applications, the issue to secure it became top urgent. In this research, a new approach to encrypt text message based on genetic algorithm operators has been proposed. The proposed approach follows a new algorithm of generating 8 bit chromosome to encrypt plain text after selecting randomly crossover point. The resulted child code is flipped by one bit using mutation operation. Two simulations are conducted to evaluate the performance of the proposed approach including execution time of encryption/decryption and throughput computations. Simulations results prove the robustness of the proposed approach to produce better performance for all evaluation metrics with res

Some problems want to be solved in image compression to make the process workable and more efficient. Much work had been done in the field of lossy image compression based on wavelet and Discrete Cosine Transform (DCT). In this paper, an efficient image compression scheme is proposed, based on a common encoding transform scheme; It consists of the following steps: 1) bi-orthogonal (tab 9/7) wavelet transform to split the image data into sub-bands, 2) DCT to de-correlate the data, 3) the combined transform stage's output is subjected to scalar quantization before being mapped to positive, 4) and LZW encoding to produce the compressed data. The peak signal-to-noise (PSNR), compression ratio (CR), and compression gain (CG) measures were used t

Exchange of information through the channels of communication can be unsafe. Communication media are not safe to send sensitive information so it is necessary to provide the protection of information from disclosure to unauthorized persons. This research presented the method to information security is done through information hiding into the cover image using a least significant bit (LSB) technique, where a text file is encrypted using a secret sharing scheme. Then, generating positions to hiding information in a random manner of cover image, which is difficult to predict hiding in the image-by-image analysis or statistical analyzes. Where it provides two levels of information security through encryption of a text file using the secret sha