Beyond the immediate content of speech, the voice can provide rich information about a speaker's demographics, including age and gender. Estimating a speaker's age and gender offers a wide range of applications, spanning from voice forensic analysis to personalized advertising, healthcare monitoring, and human-computer interaction. However, pinpointing precise age remains intricate due to age ambiguity. Specifically, utterances from individuals at adjacent ages are frequently indistinguishable. Addressing this, we propose a novel, end-to-end approach that deploys Mozilla's Common Voice dataset to transform raw audio into high-quality feature representations using Wav2Vec2.0 embeddings. These are then channeled into our self-attention-based convolutional neural network (CNN) model. To address age ambiguity, we evaluate the effects of different loss functions such as focal loss and Kullback-Leibler (KL) divergence loss. Additionally, we evaluate the accuracy of the estimation at different durations of speech. Experimental results from the Common Voice dataset underscore the efficacy of our approach, showcasing an accuracy of 87% for male speakers, 91% for female speakers and 89% overall accuracy, and an accuracy of 99.1% for gender prediction.
This research presents an experimental investigation of the rehabilitation efficiency of the damaged hybrid reinforced concrete beams with openings in the shear region. The study investigates the difference in retrofitting ability of hybrid beams compared to traditional beams and the effect of two openings compared with one opening equalized to two holes in the area. Five RC beams classified into two groups, A and B, were primarily tested to full-failure under two-point loads. The first group (A) contained beams with normal weight concrete. The second group (hybrid) included beams with lightweight concrete for web and bottom flange, whereas the top flange was made from normal concrete. Two types of openings were considered in this s
... Show MoreEx-situ bioremediation of 2,4-D herbicide-contaminated soil was studied using a slurry bioreactor operate at aerobic conditions. The performance of the slurry bioreactor was tested for three types of soil (sand, sandy loam and clay) contaminated with different concentration of 2,4-D, 200,300and500mg/kg soil. Sewage sludge was used as an inexpensive source of microorganisms which is available in large quantities in wastewater treatment plants. The results show that all biodegradation experiments demonstrated a significant decreases in 2,4-D concentration in the tested soils. The degradation efficiency in the slurry bioreactor decreases as the initial concentration of 2,4-D in the soils increases.A 100 % removal was achieved at initial con
... Show MoreA .technology analysis image using crops agricultural of grading and sorting the test to conducted was experiment The device coupling the of sensor a with camera a and 75 * 75 * 50 dimensions with shape cube studio made-factory locally the study to studio the in taken were photos and ,)blue-green - red (lighting triple with equipped was studio The .used were neural artificial and technology processing image using maturity and quality ,damage of fruits the of characteristics external value the quality 0.92062, of was value regression the damage predict to used was network neural artificial The .network the using scheme regression a of means by 0.98654 of was regression the of maturity and 0.97981 of was regression the of .algorithm Marr
... Show MoreFerritin is a key organizer of protected deregulation, particularly below risky hyperferritinemia, by straight immune-suppressive and pro-inflammatory things. , We conclude that there is a significant association between levels of ferritin and the harshness of COVID-19. In this paper we introduce a semi- parametric method for prediction by making a combination between NN and regression models. So, two methodologies are adopted, Neural Network (NN) and regression model in design the model; the data were collected from مستشفى دار التمريض الخاص for period 11/7/2021- 23/7/2021, we have 100 person, With COVID 12 Female & 38 Male out of 50, while 26 Female & 24 Male non COVID out of 50. The input variables of the NN m
... Show MoreThe Artificial Neural Network methodology is a very important & new subjects that build's the models for Analyzing, Data Evaluation, Forecasting & Controlling without depending on an old model or classic statistic method that describe the behavior of statistic phenomenon, the methodology works by simulating the data to reach a robust optimum model that represent the statistic phenomenon & we can use the model in any time & states, we used the Box-Jenkins (ARMAX) approach for comparing, in this paper depends on the received power to build a robust model for forecasting, analyzing & controlling in the sod power, the received power come from
... Show MoreThe levels of lead (pb), copper (cu), cobalt (co) and cadmium (cd) were determined in different kinds of milk and the health risks were evaluated. The mean levels were 0.73±0.21, 0.06±0.01, 0.12±0.01 and 0.14±0.01 ppm for these metals respectively. The levels of pb and cu were found to be insignificant differences (p<0.05), whereas the levels of co and cd, were no significant differences (p>0.05). The dry and liquid kinds of milk were different significantly (p<0.05), whereas the original, was no significant differences (p>0.05). The values for all metals were more than one. The metals pb and cd were detected at highest concentrations in most dry and liquid milk samples.
In the present study, a total of 245 flour samples were collected from 49 mills on both sides of Baghdad city (Al- Karkh and Al- Resafa), during the period from 1/6 - 1/12/ 2015 to detect the prolportion of iron added to the flour samples. It is found that only 45% of mills produced flour contain the prescribed percentage of iron (30-60 ppm) while 51.9% of the mills produced flour at rate is less or much more than the prescribed percentage, while only 4.1% of the mills were not added iron to the flour.
In this paper we estimate the coefficients and scale parameter in linear regression model depending on the residuals are of type 1 of extreme value distribution for the largest values . This can be regard as an improvement for the studies with the smallest values . We study two estimation methods ( OLS & MLE ) where we resort to Newton – Raphson (NR) and Fisher Scoring methods to get MLE estimate because the difficulty of using the usual approach with MLE . The relative efficiency criterion is considered beside to the statistical inference procedures for the extreme value regression model of type 1 for largest values . Confidence interval , hypothesis testing for both scale parameter and regression coefficients
... Show More