Audio-visual detection and recognition system is thought to become the most promising methods for many applications includes surveillance, speech recognition, eavesdropping devices, intelligence operations, etc. In the recent field of human recognition, the majority of the research be- coming performed presently is focused on the reidentification of various body images taken by several cameras or its focuses on recognized audio-only. However, in some cases these traditional methods can- not be useful when used alone such as in indoor surveillance systems, that are installed close to the ceiling and capture images right from above in a downwards direction and in some cases people don't look straight the cameras or it cannot be added in some area such as W.C. or sleeping room. Thus, its commonly difficult to identify any movement or breakthrough process, on the other hand when need to pursue suspect when enter a building or party to identify his location and/or listen to his speech only and isolate it from other voices or noises, the other. Hence, the use of the hybrid combination technique is very effective. In this work, we proposed a multimodal human recognition approach that utilizes both the face and audio and is based upon a deep convolutional neural network (CNN). Mainly, to solve the challenge of not capturing part of the body, final results of recognizing via separate CNNs of VGG Face16 and ResNet50 are joined together depending on the score-level combination by Weighted Sum rule to enhance recognition performance. The results show that the proposed system success to recognise each person from his voice and/or his face captured. In addition, the system can separate the person voice and isolate it from noisy environment and determine the existence of desired person.
Key-frame selection plays an important role in facial expression recognition systems. It helps in selecting the most representative frames that capture the different poses of the face. The effect of the number of selected keyframes has been studied in this paper to find its impact on the final accuracy of the emotion recognition system. Dynamic and static information is employed to select the most effective key-frames of the facial video with a short response time. Firstly, the absolute difference between the successive frames is used to reduce the number of frames and select the candidate ones which then contribute to the clustering process. The static-based information of the reduced sets of frames is then given to the fuzzy C-Means algor
... Show More
(2)
(126)
(126)
(12)
(13)
(3)
(16)
(8)
Presupposition is the background belief that is known by both the speaker and the addressee, it is tied to particular words and aspects of the surface structure that act as linguistic triggers. The present study aims at investigating whether Iraqi fourth -year university students are able to recognize the English presuppositions through the meaning of these linguistic triggers .To fulfil the basic requirements of the study, the researcher has conducted a test . The results of the study have validated the hypothesis of the work and it is found that the linguistic triggers are important tools in recognizing presuppositions.
Back ground: Diabetic nephropathy is rapidly becoming the leading cause of end-stage renal disease (ESRD). The onset and course of DN can be ameliorated to a very significant degree if intervention institutes at a point very early in the course of the development of this complication.
Objective: The aim of this study was to characterize risk factors associated with nephropathy in type I diabetes and construct a module for early prediction of diabetic nephropathy (DN) by analyzing their risk factors.
Methods: Case control design of 400 patients with type I diabetes mellitus (IDDM), aged 19-45 years. The cases were 200 diabetic patients with overt protein urea while the controls were 200 diabetic patients with no protein urea or micr
(148)
(146)
Tax state institution regards as one of the largest state institutions implementing the tax rules issuing be legislative body and achieving the goals of tax (financial, economic, social and political). So, the tax management should pay attention to the procedures enabling it to achieve those rules starting from the procedures of tax restrict and ending by tax allocation where the process of assessment the taxation must relaxing on modern methods. The problem of the study raising from that in spite of there is a low obliging the taxable person (registered or not) to submit a declaration about his income and the achieved profit to be the base of taxation˒ where the other ways are secondary ways helping in rejection of t
... Show More
Surface electromyography (sEMG) and accelerometer (Acc) signals play crucial roles in controlling prosthetic and upper limb orthotic devices, as well as in assessing electrical muscle activity for various biomedical engineering and rehabilitation applications. In this study, an advanced discrimination system is proposed for the identification of seven distinct shoulder girdle motions, aimed at improving prosthesis control. Feature extraction from Time-Dependent Power Spectrum Descriptors (TDPSD) is employed to enhance motion recognition. Subsequently, the Spectral Regression (SR) method is utilized to reduce the dimensionality of the extracted features. A comparative analysis is conducted between the Linear Discriminant Analysis (LDA) class
... Show More
(1)
(1)