scholarly journals Benchmarking deep network architectures for ethnicity recognition using a new large face dataset

2020 ◽  
Vol 31 (7-8) ◽  
Author(s):  
Antonio Greco ◽  
Gennaro Percannella ◽  
Mario Vento ◽  
Vincenzo Vigilante
Keyword(s):  
Neural Networks ◽  
Network Architectures ◽  
Related Factors ◽  
Deep Network ◽  
Soft Biometrics ◽  
Face Images ◽  
Race Effect ◽  
Effective Training ◽  
The One ◽  
Better Than

Abstract Although in recent years we have witnessed an explosion of the scientific research in the recognition of facial soft biometrics such as gender, age and expression with deep neural networks, the recognition of ethnicity has not received the same attention from the scientific community. The growth of this field is hindered by two related factors: on the one hand, the absence of a dataset sufficiently large and representative does not allow an effective training of convolutional neural networks for the recognition of ethnicity; on the other hand, the collection of new ethnicity datasets is far from simple and must be carried out manually by humans trained to recognize the basic ethnicity groups using the somatic facial features. To fill this gap in the facial soft biometrics analysis, we propose the VGGFace2 Mivia Ethnicity Recognition (VMER) dataset, composed by more than 3,000,000 face images annotated with 4 ethnicity categories, namely African American, East Asian, Caucasian Latin and Asian Indian. The final annotations are obtained with a protocol which requires the opinion of three people belonging to different ethnicities, in order to avoid the bias introduced by the well-known other race effect. In addition, we carry out a comprehensive performance analysis of popular deep network architectures, namely VGG-16, VGG-Face, ResNet-50 and MobileNet v2. Finally, we perform a cross-dataset evaluation to demonstrate that the deep network architectures trained with VMER generalize on different test sets better than the same models trained on the largest ethnicity dataset available so far. The ethnicity labels of the VMER dataset and the code used for the experiments are available upon request at https://mivia.unisa.it.

Infant Cry Recognition System

2019 ◽  
Vol 11 (1) ◽  
pp. 15-32
Author(s):  
Yosra Abdulaziz Mohammed
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Markov Models ◽  
Hidden Markov ◽  
Recognition System ◽  
Acoustic Features ◽  
Infant Cry ◽  
Identification Rate ◽  
The One ◽  
Better Than

Cries of infants can be seen as an indicator of pain. It has been proven that crying caused by pain, hunger, fear, stress, etc., show different cry patterns. The work presented here introduces a comparative study between the performance of two different classification techniques implemented in an automatic classification system for identifying two types of infants' cries, pain, and non-pain. The techniques are namely, Continuous Hidden Markov Models (CHMM) and Artificial Neural Networks (ANN). Two different sets of acoustic features were extracted from the cry samples, those are MFCC and LPCC, the feature vectors generated by each were eventually fed into the classification module for the purpose of training and testing. The results of this work showed that the system based on CDHMM have better performance than that based on ANN. CDHMM gives the best identification rate at 96.1%, which is much higher than 79% of ANN whereby in general the system based on MFCC features performed better than the one that utilizes LPCC features.

NEURAL NETWORKS IN PETROLEUM ENGINEERING: A CASE STUDY

1996 ◽  
Vol 07 (02) ◽  
pp. 187-194 ◽  
Author(s):  
M.A. VUKELIC ◽  
E.N. MIRANDA
Keyword(s):  
Neural Networks ◽  
Test Area ◽  
Petroleum Engineering ◽  
Oil Reservoir ◽  
Network Architectures ◽  
Generalization Capability ◽  
Learning Capability ◽  
Hidden Neurons ◽  
Better Than

A multilayer neural network has been used for deciding which oil reservoir layer has to be perforated. Many network architectures were tested until we found those with the best generalization capability. The network performs better than human experts and its achievements are higher than the historical average in the test area. As in other applications of neural networks, the learning capability improves with more hidden neurons but the generalization does not.

scholarly journals Effective training of convolutional neural networks for age estimation based on knowledge distillation

2021 ◽  
Author(s):  
Antonio Greco ◽  
Alessia Saggese ◽  
Mario Vento ◽  
Vincenzo Vigilante
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Age Estimation ◽  
Resource Constraints ◽  
Training Procedure ◽  
Face Images ◽  
Effective Training ◽  
Knowledge Distillation ◽  
Student Models ◽  
Teacher Model

AbstractAge estimation from face images can be profitably employed in several applications, ranging from digital signage to social robotics, from business intelligence to access control. Only in recent years, the advent of deep learning allowed for the design of extremely accurate methods based on convolutional neural networks (CNNs) that achieve a remarkable performance in various face analysis tasks. However, these networks are not always applicable in real scenarios, due to both time and resource constraints that the most accurate approaches often do not meet. Moreover, in case of age estimation, there is the lack of a large and reliably annotated dataset for training deep neural networks. Within this context, we propose in this paper an effective training procedure of CNNs for age estimation based on knowledge distillation, able to allow smaller and simpler “student” models to be trained to match the predictions of a larger “teacher” model. We experimentally show that such student models are able to almost reach the performance of the teacher, obtaining high accuracy over the LFW+, LAP 2016 and Adience datasets, but being up to 15 times faster. Furthermore, we evaluate the performance of the student models in the presence of image corruptions, and we demonstrate that some of them are even more resilient to these corruptions than the teacher model.

scholarly journals Gender recognition in the wild: a robustness evaluation over corrupted images

2020 ◽  
Author(s):  
Antonio Greco ◽  
Alessia Saggese ◽  
Mario Vento ◽  
Vincenzo Vigilante
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Network Architectures ◽  
Gender Recognition ◽  
Face Images ◽  
The Face ◽  
In The Wild ◽  
The Neural Networks ◽  
The Common ◽  
And Gender

AbstractIn the era of deep learning, the methods for gender recognition from face images achieve remarkable performance over most of the standard datasets. However, the common experimental analyses do not take into account that the face images given as input to the neural networks are often affected by strong corruptions not always represented in standard datasets. In this paper, we propose an experimental framework for gender recognition “in the wild”. We produce a corrupted version of the popular LFW+ and GENDER-FERET datasets, that we call LFW+C and GENDER-FERET-C, and evaluate the accuracy of nine different network architectures in presence of specific, suitably designed, corruptions; in addition, we perform an experiment on the MIVIA-Gender dataset, recorded in real environments, to analyze the effects of mixed image corruptions happening in the wild. The experimental analysis demonstrates that the robustness of the considered methods can be further improved, since all of them are affected by a performance drop on images collected in the wild or manually corrupted. Starting from the experimental results, we are able to provide useful insights for choosing the best currently available architecture in specific real conditions. The proposed experimental framework, whose code is publicly available, is general enough to be applicable also on different datasets; thus, it can act as a forerunner for future investigations.

Obstacles to Claiming Permanence and Injury-Relatedness for “Posttraumatic” Headache

2009 ◽  
Vol 14 (3) ◽  
pp. 3-6
Author(s):  
Robert J. Barth
Keyword(s):  
Complete Recovery ◽  
World Health ◽  
Related Factors ◽  
Posttraumatic Headache ◽  
Sixth Edition ◽  
Causative Relationship ◽  
Short Period ◽  
The One ◽  
Health Organization ◽  
Subjective Reports

Abstract “Posttraumatic” headaches claims are controversial because they are subjective reports often provided in the complex of litigation, and the underlying pathogenesis is not defined. This article reviews principles and scientific considerations in the AMAGuides to the Evaluation of Permanent Impairment (AMA Guides) that should be noted by evaluators who examine such cases. Some examples in the AMA Guides, Sixth Edition, may seem to imply that mild head trauma can cause permanent impairment due to headache. The author examines scientific findings that present obstacles to claiming that concussion or mild traumatic brain injury is a cause of permanent headache. The World Health Organization, for example, found a favorable prognosis for posttraumatic headache, and complete recovery over a short period of time was the norm. Other studies have highlighted the lack of a dose-response correlation between trauma and prolonged headache complaints, both in terms of the frequency and the severity of trauma. On the one hand, scientific studies have failed to support the hypothesis of a causative relationship between trauma and permanent or prolonged headaches; on the other hand, non–trauma-related factors are strongly associated with complaints of prolonged headache.

Subjective Time Preference and Willingness to Pay for an Energy-Saving Durable Good

2001 ◽  
Vol 32 (3) ◽  
pp. 133-141 ◽  
Author(s):  
Gerrit Antonides ◽  
Sophia R. Wunderink
Keyword(s):  
Willingness To Pay ◽  
Energy Saving ◽  
Subjective Time ◽  
Durable Good ◽  
Discount Function ◽  
Hyperbolic Discount ◽  
The One ◽  
Two Parameter ◽  
Difference Effect ◽  
Better Than

Summary: Different shapes of individual subjective discount functions were compared using real measures of willingness to accept future monetary outcomes in an experiment. The two-parameter hyperbolic discount function described the data better than three alternative one-parameter discount functions. However, the hyperbolic discount functions did not explain the common difference effect better than the classical discount function. Discount functions were also estimated from survey data of Dutch households who reported their willingness to postpone positive and negative amounts. Future positive amounts were discounted more than future negative amounts and smaller amounts were discounted more than larger amounts. Furthermore, younger people discounted more than older people. Finally, discount functions were used in explaining consumers' willingness to pay for an energy-saving durable good. In this case, the two-parameter discount model could not be estimated and the one-parameter models did not differ significantly in explaining the data.

Exploring the Multidimensional Facets of Authoritarianism: Authoritarian Aggression and Social Dominance Orientation

2008 ◽  
Vol 67 (1) ◽  
pp. 51-60 ◽  
Author(s):  
Stefano Passini
Keyword(s):  
Social Dominance ◽  
Factor Model ◽  
Three Dimensional ◽  
Social Dominance Orientation ◽  
Model Fit ◽  
Regression Analyses ◽  
One Dimensional ◽  
Confirmatory Factor ◽  
The One ◽  
Better Than

The relation between authoritarianism and social dominance orientation was analyzed, with authoritarianism measured using a three-dimensional scale. The implicit multidimensional structure (authoritarian submission, conventionalism, authoritarian aggression) of Altemeyer’s (1981, 1988) conceptualization of authoritarianism is inconsistent with its one-dimensional methodological operationalization. The dimensionality of authoritarianism was investigated using confirmatory factor analysis in a sample of 713 university students. As hypothesized, the three-factor model fit the data significantly better than the one-factor model. Regression analyses revealed that only authoritarian aggression was related to social dominance orientation. That is, only intolerance of deviance was related to high social dominance, whereas submissiveness was not.

Color Space Transformation using Neural Networks

2019 ◽  
Vol 2019 (1) ◽  
pp. 153-158
Author(s):  
Lindsay MacDonald
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Color Space ◽  
Reflectance Spectra ◽  
Network Architectures ◽  
Color Spaces ◽  
Natural Materials ◽  
Space Transformation ◽  
Color Space Transformation

We investigated how well a multilayer neural network could implement the mapping between two trichromatic color spaces, specifically from camera R,G,B to tristimulus X,Y,Z. For training the network, a set of 800,000 synthetic reflectance spectra was generated. For testing the network, a set of 8,714 real reflectance spectra was collated from instrumental measurements on textiles, paints and natural materials. Various network architectures were tested, with both linear and sigmoidal activations. Results show that over 85% of all test samples had color errors of less than 1.0 ΔE2000 units, much more accurate than could be achieved by regression.

Designing Women

2018 ◽  
Author(s):  
J. E. Smyth
Keyword(s):  
Working Women ◽  
Costume Design ◽  
Creative Force ◽  
Women's Fashion ◽  
The One ◽  
Studio Era ◽  
Average Woman ◽  
Better Than

During the early 1940s, journalists observed that after years of men controlling women’s fashion, Hollywood had become “a fashion center in which women designers are getting to be a big power.” In a town where “the working girl is queen,” it was women who really knew how to dress working women. Edith Head’s name dominates Hollywood costume design. Though a relatively poor sketch artist who refused to sew in public, Head understood what the average woman wanted to wear and knew better than anyone how to craft her image as the-one-and-only Edith Head. However, she was one of many women who designed Hollywood glamour in the studio era. This chapter juxtaposes Head’s career with that of a younger, fiercely independent designer who would quickly upstage Head as a creative force. In many senses, Dorothy Jeakins’s postwar career ascent indicated the waning of the Hollywood system and the powerful relationship between female designers, stars, and fans.

scholarly journals Do 3D Face Images Capture Cues of Strength, Weight, and Height Better than 2D Face Images do?

2021 ◽  
Vol 7 (3) ◽  
pp. 209-219
Author(s):  
Iris J Holzleitner ◽  
Alex L Jones ◽  
Kieran J O’Shea ◽  
Rachel Cassar ◽  
Vanessa Fasolt ◽  
...  
Keyword(s):  
Three Dimensional ◽  
Physical Characteristics ◽  
Two Dimensional ◽  
3D Images ◽  
3D Face ◽  
Face Images ◽  
Similar Accuracy ◽  
2D And 3D ◽  
2D Images ◽  
Better Than

Abstract Objectives A large literature exists investigating the extent to which physical characteristics (e.g., strength, weight, and height) can be accurately assessed from face images. While most of these studies have employed two-dimensional (2D) face images as stimuli, some recent studies have used three-dimensional (3D) face images because they may contain cues not visible in 2D face images. As equipment required for 3D face images is considerably more expensive than that required for 2D face images, we here investigated how perceptual ratings of physical characteristics from 2D and 3D face images compare. Methods We tested whether 3D face images capture cues of strength, weight, and height better than 2D face images do by directly comparing the accuracy of strength, weight, and height ratings of 182 2D and 3D face images taken simultaneously. Strength, height and weight were rated by 66, 59 and 52 raters respectively, who viewed both 2D and 3D images. Results In line with previous studies, we found that weight and height can be judged somewhat accurately from faces; contrary to previous research, we found that people were relatively inaccurate at assessing strength. We found no evidence that physical characteristics could be judged more accurately from 3D than 2D images. Conclusion Our results suggest physical characteristics are perceived with similar accuracy from 2D and 3D face images. They also suggest that the substantial costs associated with collecting 3D face scans may not be justified for research on the accuracy of facial judgments of physical characteristics.

Sign in / Sign up

Export Citation Format

Share Document