Literature Review on Gender Prediction Model using CNN Algorithm
0 likes52 views
The document discusses a literature review on gender prediction models using CNN algorithms. It summarizes previous research that used techniques like CNNs, K-means clustering, and open-CV to develop models for gender classification from images with over 90% accuracy. The paper also proposes developing an enhanced gender prediction model using a CNN approach for pre-processing images and evaluating the accuracy of the model. It suggests CNNs can effectively extract features from faces and classify gender by training on labeled image datasets. Evaluation of the proposed model showed 98.7% accuracy on open-CV and 94% on CNN datasets.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 316
2. Clarification of Prediction Model
To locate a human face, the system needs to capture an
image using a camera and a frame-grabber to process the
image, search the image for important features and then use
these features to determine the location of the face.
Spontaneous predicting his/her image and preprocessing
the image, checking the accuracy of the model.
Convolutional neural network(CNN)that uses a high recall
fast face detector for generatingregionproposalstoimprove
the speed of the algorithm. Considerable experiments show
the model or system able to collect local and global data
related to facial images and perform significantlybetterthan
a competitive algorithm or method four tasks. Age and
gender, two important facial attributes, play a very
infrastructure role in social interactions, making age and
gender approximation from a face image is an important
task in applications,suchasaccesscontrol,human-computer
interaction, law enforcement, marketing intelligence, and
visual surveillance, etc. we will be doing face recognition,
face prediction and, we will be using CNN (Convolutional
Neural Networks) for age and gender predictions.
3. LITERATURE REVIEW
In this section, we briefly review the age and gender
classification literature and describe both theearlymethods
The main aim of this method fixing age and gender
classification and check the accuracy of the model for his or
her images.
The paper gives you information about the technology used
in the gender detection model.
The models using an algorithm for detecting an image pose
prediction and recognition of his or her images using
convolutional neural network algorithm it enhances
performance and high face detector for improving thespeed
of the model significantly better in performance .and
performing many more tasks[1].
We have the prospect to learn and classification methodand
using a convolutional neural network technique to enhance
the performance can be attained gender classification task
that we decide for working on the convolutional neural
network[2]. using VGGnet architecture which can be used in
an extreme level of training data for CNN-based algorithms
we supposed our work on current unfiltered pictures for his
or her gender prediction.
Fig:- Graph of Accuracy
This paper introduces an approach of gender detection ofa
person using facial images using a techniquefordata mining.
prediction of gender is a very useful technique in human-
computer Interaction and classification of the model or a
system, classification is a powerful technique used in
categorical data, classify males and females also using a
functional tree, J48, and other is used for checking a
maximum accuracy of a model[3]. The model used in this
paper for attaining an accuracy level of 95.66%
This research suggests that the new algorithm for
automatic live gender recognition using a support vector
machine is used for the classification of the model. The
implementation of work results on FEI, live images, and
detection accuracy is 97% in the FEI dataset, 95% in own
dataset this proposed methodology is compared to the
previous method for better prediction and which will be
helpful in the real-time system[4].
Fig:- Working Model
Finally, we look over the accuracy of the classification
dataset for age and gender our method enhances
performance in both age and his or her prediction and also
classification of gender, significantly exceeding performing
the models. For future enhancement of work support. then
we will use a Convolutional Neural Network algorithm for
age estimation and prediction of gender.](https://image.slidesharecdn.com/irjet-v9i356-220915115955-17a37a76/85/Literature-Review-on-Gender-Prediction-Model-using-CNN-Algorithm-2-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 317
Previously growing technology regarding the security-
related problem in our daily life. In this paper, we discuss
biometric traits in the gender prediction model forhisorher
identification and also used for reducing search space tests
[5]. In this model using 20 subsequent layers they have a
different window size of Convolutional Neural
Network(CNN) and the layer are fully connected layers for
extraction and classification purposes.Using biometrictraits
takes an input of face, fingerprint, eyes images classification
of the SDMULA-HMTand KVK-Multimodal datasetsandgives
an output using multimodal biometrics data fusion schemes
to increase the accuracy of the gender prediction model.
Prediction of gender and age of classifying unconstrained
facial images. This research area is useful in the real-world
application or real-world area traditional methods not
providing security in real-world application lots of security
issues are arises[6]. Now this model uses Convolutional
Neural Network (CNN) it is commonly used forclassification
purposes and also performing classificationtasks duetoface
images or analysis of a face In This model performs an end-
to-end Convolutional Neural Network approach to
accomplish gender classification and robust group of age.
The automatic age and prediction of gender extracting from
human face images. The Convolutional Neural Network
approach is used to accomplish a task for predicting age and
gender prediction. In this model usingvarioushardwareand
software are used to increase’s the accuracy of themodel for
predicting age and gender (his/her)andusedinonlinesocial
networking websites, social networks. In this, we use a
classification method and Convolutional Neural Network
technique [7].
That Deep Age Distribution Learning (DADL) in this
estimating age and generating Gaussian age. taking multiple
face images with labeled finding age using a standard
deviation. Give a mean age and standard deviation and find
out a Gaussian age distribution for each face image using a
target training dataset. First, detect the region of face and
alignment of face[8]. Then using a Convolutional Neural
Network for training based on VGG face and predicting an
age distribution using a trained age dataset. Finally using
Ensemble method for combining multiple models and
enhance the accuracy of the model.
The Gender prediction from social media comments with
artificial Intelligence In this used artificial Intelligence for
predicting an age group and also used a machine learning
technique[9]. Nowadays many companiesand organizations
delivered a product and provide a service to their customer
via an online platform but not all customers are the same or
have different interests. Gender is the main reason behind
these problems. If Gender is determined correctly on an
online platform or social media then the amount or cost of
the product increases usinga machinelearningtechniquefor
analyzing the comments ofcompanies. Predictingthegender
enhances the accuracy level of the model.
4. DISCUSSION
We use our method for classification dataset accuracy for
the prediction of gender and age. also predicting an age
group. Classifying the correct gender of a person. containing
dataset for labels of gender like male and female.
Using a Deep Convolutional Neural Network (D-CNN) for
recognition and predicting of an object. In modern
Convolutional Neural Network(CNN) like VGGNet, LeNet-5
network architecture is completely moderate becauseofthe
limited number of resources such as time, memory,
processing power, and another algorithm for challenging
huge networks. The aim is to learn to preprocess the input
image and convert it into meaningful floating-point tensors.
A neural network with a huge number of neurons layer is
there, the have vogue because of unpredictedrisein both the
computational power with the use of Graphical processing
units and dataset easily available on the Internet. We are
going to collect a dataset and also used a dataset for our
experiment. It contains 100k celebrity face images for
reading images and convert into Gray-Scaleforthedetection
or predicting cropping of visible faces using Convolutional
Neural Network (CNN)
It measures the percentage of face images that were
classified into the correct age group and gender :
Accuracy = Accurate prediction
Total prediction
5. CONCLUSION
We initiated the classification of gender(his or her image)
and age group unfiltered real-worldfacial images.Wepasses
a task as a multiclass classification of the problem, trained
the model or a system according to prediction, and trained
the model for the accomplishment of our result. our
approach achieves the best result not only for gender but
also on age estimation. We trained our network model to
classify facial images into eight age groups and enhance the
performance or increase the accuracy of the model. Our aim
for proposed a model that is originally trained on age and
gender prediction on a large data scale. The image
preprocessing algorithm handles some of the variability
observed in the real world faces. And this confirms the
model relevancy forgenderclassificationandageestimation.
For future works, we will consider a Convolutional Neural
Network(CNN) architecture for preprocessing animageand
age prediction. Also age prediction or estimating the ageofa
human’s face will be interesting to explore in the future.
6. REFERENCES
[1] D. T. Ramotsoela, G. P. Hancke, and A. M. Abu-Mahfouz,
Hum.-Centric Comput. Inf. Sci., vol. 9, pp. 13, Apr. 2019.](https://image.slidesharecdn.com/irjet-v9i356-220915115955-17a37a76/85/Literature-Review-on-Gender-Prediction-Model-using-CNN-Algorithm-3-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 318
[2] L. Srinivasan and C. Nalini, Cluster Comput., vol. 22, no.
S5, pp. 12101-12110, Sep. 2019.
[3] Beech JR, Mackintosh IC (2005) Do differences in sex
hormones affect, Evidence from digit ratio and sex-role
identity as determinants of the sex of handwriting. Pers
Individ Dif 39(2):459–468.
[4] R. Rothe, R. Timofte and L. V. Gool," in Proc. IEEE
International Conference on Computer Vision Workshops,
Dec 2018.
[5]O. M. Parkhi, A. Vedaldi, and A. Zisserman, "Face
recognition," in Proc. Machine Vision Vol. 1, No 3, pp. 6-17,
2015.
[6] J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei,
"ImageNet: A large-scale hierarchical image database",Proc.
IEEE Conf. Comput. Vis.PatternRecognit.,pp.248-255,2009.
[7] P. Felzenszwalb, R. Girshick, D. McAllester ,D. Ramanan,
"Object detection with discriminatively trained part-based
models", IEEE Trans. Pattern Anal.Mach.Intell.,vol.32,no.9,
pp. 1627-1645, Sep. 2010.
[8] Comparison of CNN Algorithms on Hyperspectral Image
Classification in Agricultural LandsbyTien-HengHsieh†and
Jen-Fu Kiang.
[9] Lekha Athota Student, Bachelor of Science Information
Technology,VinodKumarShukla DepartmentofEngineering
and Architecture, Nitin Pandey Amity Institute of
Information Technology, Amity University June 4-5, 2020.
[10] Nadu International Journal for Technological Research
InEngineering Volume 6, Issue 7, March-2019](https://image.slidesharecdn.com/irjet-v9i356-220915115955-17a37a76/85/Literature-Review-on-Gender-Prediction-Model-using-CNN-Algorithm-4-320.jpg)
Literature Review on Gender Prediction Model using CNN Algorithm
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 315 Literature Review on Gender Prediction Model using CNN Algorithm Kamekshi Gaur1, Dr. Pawan Kumar2 1MCA, School of CS & IT, Jain University, Bangalore, India. 2Professor, School of CS & IT, Jain University, Bangalore, India. ---------------------------------------------------------------------***---------------------------------------------------------------------- Abstract – In Gender Prediction model using the CNN Algorithm to the establishment of an Automatic gender detection has now pertinent to a supplement of its using hardware and software, because of the growth of the online social networks, malls. Nowadays prediction of gender is a very challenging task in real-time application. Prediction of a gender enhancement in future growth using a CNN(Convolutional Neural Network) for preprocessing the images and checking the accuracy of a model. CNN(ConvolutionalNeuralNetwork)architectureincludes the derivation of facial and classification of the model. The CNN algorithm purpose how to preprocess the input image and converting into a floating-point tensor delivered into Convolutional Neural Network. For knowledge, purposes tensors are used to store data they can define a multidimensional array. The tested result, analyzing for accuracy of classification or checking the accuracy result for predicting his/her image. The tested result show accuracy which is 98.7% using the open-cv dataset and 94% accuracy using CNN dataset. Key Words: CNN -algorithm, K-means clustering algorithm, machine learning, Open-CV, recommendation system, grayscale detection. 1. INTRODUCTION Gender detection plays a significant role in modern technology. detection of gender has vast dynamic applicationslikesocialinteraction,securitymaintenance,and surveillance, video games, human-computer interaction, criminal identification, mobile application, commercial development, monitoring application, etc. It has occupied a great space in the field of facial recognition. The main purpose of gender detection is to differentiate males and females based on the different features of humans. In this paper, wehave used Open-CV andCNN algorithm K-means clustering algorithm for gender recognition to extemporize the previously used method and to obtain an accurateresult. depicts how to processfaceimagethroughK- means clustering algorithm and find the pattern, extract feature to recognize gender from image accurately. The performance of the K-means clustering algorithm depends upon the highly efficient clusters that it forms. But choosing the optimal number of clusters is a big task. accommodate to the real-time of gender prediction, it is crucial to improve the algorithms from ever so often to achieve high accuracy levels and build more efficient and accurate systems. Based on the features used extracting the facial images approaches can be divided into 2 different categories: geometric feature-based methods and appearance-based methods. Configuration refers to the distance between various facial image features such as eyes, nose, chin, and lips. Facial features can be extracted from the facial image using a k-means algorithm that returns the coordinates of various features. In current works for classification gender and age group an inspiring sign of performance in machine learning and Convolutional NeuralNetwork. The aim of this paper is end-to-endlearning classification of a model andpredict his/herimageandcheck the accuracy level of the model. In Convolutional Neural Network(CNN) algorithm : 1. reading the images. 2. Decode JPEG content into an RGB grid of pixels. 3. converting into Floating points tensors taking input to neural network 4. Rescale the pixels values between 0 to 255. Trained the neural network efficiently or effectively. Fig:- Flow Chat
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 316 2. Clarification of Prediction Model To locate a human face, the system needs to capture an image using a camera and a frame-grabber to process the image, search the image for important features and then use these features to determine the location of the face. Spontaneous predicting his/her image and preprocessing the image, checking the accuracy of the model. Convolutional neural network(CNN)that uses a high recall fast face detector for generatingregionproposalstoimprove the speed of the algorithm. Considerable experiments show the model or system able to collect local and global data related to facial images and perform significantlybetterthan a competitive algorithm or method four tasks. Age and gender, two important facial attributes, play a very infrastructure role in social interactions, making age and gender approximation from a face image is an important task in applications,suchasaccesscontrol,human-computer interaction, law enforcement, marketing intelligence, and visual surveillance, etc. we will be doing face recognition, face prediction and, we will be using CNN (Convolutional Neural Networks) for age and gender predictions. 3. LITERATURE REVIEW In this section, we briefly review the age and gender classification literature and describe both theearlymethods The main aim of this method fixing age and gender classification and check the accuracy of the model for his or her images. The paper gives you information about the technology used in the gender detection model. The models using an algorithm for detecting an image pose prediction and recognition of his or her images using convolutional neural network algorithm it enhances performance and high face detector for improving thespeed of the model significantly better in performance .and performing many more tasks[1]. We have the prospect to learn and classification methodand using a convolutional neural network technique to enhance the performance can be attained gender classification task that we decide for working on the convolutional neural network[2]. using VGGnet architecture which can be used in an extreme level of training data for CNN-based algorithms we supposed our work on current unfiltered pictures for his or her gender prediction. Fig:- Graph of Accuracy This paper introduces an approach of gender detection ofa person using facial images using a techniquefordata mining. prediction of gender is a very useful technique in human- computer Interaction and classification of the model or a system, classification is a powerful technique used in categorical data, classify males and females also using a functional tree, J48, and other is used for checking a maximum accuracy of a model[3]. The model used in this paper for attaining an accuracy level of 95.66% This research suggests that the new algorithm for automatic live gender recognition using a support vector machine is used for the classification of the model. The implementation of work results on FEI, live images, and detection accuracy is 97% in the FEI dataset, 95% in own dataset this proposed methodology is compared to the previous method for better prediction and which will be helpful in the real-time system[4]. Fig:- Working Model Finally, we look over the accuracy of the classification dataset for age and gender our method enhances performance in both age and his or her prediction and also classification of gender, significantly exceeding performing the models. For future enhancement of work support. then we will use a Convolutional Neural Network algorithm for age estimation and prediction of gender.
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 317 Previously growing technology regarding the security- related problem in our daily life. In this paper, we discuss biometric traits in the gender prediction model forhisorher identification and also used for reducing search space tests [5]. In this model using 20 subsequent layers they have a different window size of Convolutional Neural Network(CNN) and the layer are fully connected layers for extraction and classification purposes.Using biometrictraits takes an input of face, fingerprint, eyes images classification of the SDMULA-HMTand KVK-Multimodal datasetsandgives an output using multimodal biometrics data fusion schemes to increase the accuracy of the gender prediction model. Prediction of gender and age of classifying unconstrained facial images. This research area is useful in the real-world application or real-world area traditional methods not providing security in real-world application lots of security issues are arises[6]. Now this model uses Convolutional Neural Network (CNN) it is commonly used forclassification purposes and also performing classificationtasks duetoface images or analysis of a face In This model performs an end- to-end Convolutional Neural Network approach to accomplish gender classification and robust group of age. The automatic age and prediction of gender extracting from human face images. The Convolutional Neural Network approach is used to accomplish a task for predicting age and gender prediction. In this model usingvarioushardwareand software are used to increase’s the accuracy of themodel for predicting age and gender (his/her)andusedinonlinesocial networking websites, social networks. In this, we use a classification method and Convolutional Neural Network technique [7]. That Deep Age Distribution Learning (DADL) in this estimating age and generating Gaussian age. taking multiple face images with labeled finding age using a standard deviation. Give a mean age and standard deviation and find out a Gaussian age distribution for each face image using a target training dataset. First, detect the region of face and alignment of face[8]. Then using a Convolutional Neural Network for training based on VGG face and predicting an age distribution using a trained age dataset. Finally using Ensemble method for combining multiple models and enhance the accuracy of the model. The Gender prediction from social media comments with artificial Intelligence In this used artificial Intelligence for predicting an age group and also used a machine learning technique[9]. Nowadays many companiesand organizations delivered a product and provide a service to their customer via an online platform but not all customers are the same or have different interests. Gender is the main reason behind these problems. If Gender is determined correctly on an online platform or social media then the amount or cost of the product increases usinga machinelearningtechniquefor analyzing the comments ofcompanies. Predictingthegender enhances the accuracy level of the model. 4. DISCUSSION We use our method for classification dataset accuracy for the prediction of gender and age. also predicting an age group. Classifying the correct gender of a person. containing dataset for labels of gender like male and female. Using a Deep Convolutional Neural Network (D-CNN) for recognition and predicting of an object. In modern Convolutional Neural Network(CNN) like VGGNet, LeNet-5 network architecture is completely moderate becauseofthe limited number of resources such as time, memory, processing power, and another algorithm for challenging huge networks. The aim is to learn to preprocess the input image and convert it into meaningful floating-point tensors. A neural network with a huge number of neurons layer is there, the have vogue because of unpredictedrisein both the computational power with the use of Graphical processing units and dataset easily available on the Internet. We are going to collect a dataset and also used a dataset for our experiment. It contains 100k celebrity face images for reading images and convert into Gray-Scaleforthedetection or predicting cropping of visible faces using Convolutional Neural Network (CNN) It measures the percentage of face images that were classified into the correct age group and gender : Accuracy = Accurate prediction Total prediction 5. CONCLUSION We initiated the classification of gender(his or her image) and age group unfiltered real-worldfacial images.Wepasses a task as a multiclass classification of the problem, trained the model or a system according to prediction, and trained the model for the accomplishment of our result. our approach achieves the best result not only for gender but also on age estimation. We trained our network model to classify facial images into eight age groups and enhance the performance or increase the accuracy of the model. Our aim for proposed a model that is originally trained on age and gender prediction on a large data scale. The image preprocessing algorithm handles some of the variability observed in the real world faces. And this confirms the model relevancy forgenderclassificationandageestimation. For future works, we will consider a Convolutional Neural Network(CNN) architecture for preprocessing animageand age prediction. Also age prediction or estimating the ageofa human’s face will be interesting to explore in the future. 6. REFERENCES [1] D. T. Ramotsoela, G. P. Hancke, and A. M. Abu-Mahfouz, Hum.-Centric Comput. Inf. Sci., vol. 9, pp. 13, Apr. 2019.
- 4. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 318 [2] L. Srinivasan and C. Nalini, Cluster Comput., vol. 22, no. S5, pp. 12101-12110, Sep. 2019. [3] Beech JR, Mackintosh IC (2005) Do differences in sex hormones affect, Evidence from digit ratio and sex-role identity as determinants of the sex of handwriting. Pers Individ Dif 39(2):459–468. [4] R. Rothe, R. Timofte and L. V. Gool," in Proc. IEEE International Conference on Computer Vision Workshops, Dec 2018. [5]O. M. Parkhi, A. Vedaldi, and A. Zisserman, "Face recognition," in Proc. Machine Vision Vol. 1, No 3, pp. 6-17, 2015. [6] J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, "ImageNet: A large-scale hierarchical image database",Proc. IEEE Conf. Comput. Vis.PatternRecognit.,pp.248-255,2009. [7] P. Felzenszwalb, R. Girshick, D. McAllester ,D. Ramanan, "Object detection with discriminatively trained part-based models", IEEE Trans. Pattern Anal.Mach.Intell.,vol.32,no.9, pp. 1627-1645, Sep. 2010. [8] Comparison of CNN Algorithms on Hyperspectral Image Classification in Agricultural LandsbyTien-HengHsieh†and Jen-Fu Kiang. [9] Lekha Athota Student, Bachelor of Science Information Technology,VinodKumarShukla DepartmentofEngineering and Architecture, Nitin Pandey Amity Institute of Information Technology, Amity University June 4-5, 2020. [10] Nadu International Journal for Technological Research InEngineering Volume 6, Issue 7, March-2019