Gangesh Trivedi

Bio: Gangesh Trivedi is an academic researcher. The author has contributed to research in topics: Artificial neural network. The author has an hindex of 1, co-authored 1 publications receiving 4 citations.

Gender Classification and Age Estimation using Neural Networks: A Survey

Abstract: Researchers have shown more interest in soft biometrics area to fill the commination gaps between humans and machines with the growth of real-world application has increased day to day life. Soft-biometric consists of age, gender, ethnicity, height, facial measurements and etc. This paper contains a detail discussion about the contribution of the researchers in the area of gender classification and age estimation using neural networking. Most of the work is done using Convolutional neural networks and auto encoders. Various elements related to neural network model such as dataset, findings, calculative metrics and results are embraced for effortless interpretation of tabular correlation research. Finally, the authors summarize germane tasks for future various research aspects.

Klasifikasi Gender dan Usia Berdasarkan Citra Wajah Manusia Menggunakan Convolutional Neural Network

Abstract: Fokus pada studi ini adalah untuk membuat sebuah model klasifikasi untuk memprediksi gender dan usia menggunakan citra wajah manusia. Metode klasifikasi studi ini menggunakan convolutional neural network. Tujuan dari studi ini adalah membangun model klasifikasi yang dapat memprediksi gender dan usia dari data citra wajah manusia yang sudah ada. Untuk melakukannya, terdapat beberapa proses yang dilakukan, mulai dari pengumpulan data, praproses, pembagian data, pelatihan dan pengujian data. Data yang digunakan berasal dari UTKFace yang memiliki 23.708 data citra wajah manusia dengan dua label, yaitu gender dan usia. Label gender terdiri dari laki- laki dan perempuan, serta label usia terdiri dari usia 0-20, 21-40, dan 41+ tahun. Penelitian ini menyimpulkan bahwa metode convolutional neural network dapat mengklasifikasi gender dengan performa akurasi 89,18%, presisi 89,28%, dan sensitivitas 89,17%. Dan dalam mengklasifikasi usia dapat menghasilkan performa akurasi 74,14%, presisi 78,07%, dan sensitivitas 70,65%. Kata kunci: Gender, Usia, CNN

Performance of Deep Learning Inception Model and MobileNet Model on Gender Prediction Through Eye Image

Abstract: Convolutional neural network (CNN) is one of the neural networks used in image data. CNN has a good ability to detect objects in an image. This study discusses the comparison of two deep learning models based on convolutional neural network, namely the Inception-V3 method and the MobileNet method. Both algorithms are analyzed fairly on gender classification using eye images. There have been many research completions that have conducted studies on gender classification based on faces, but gender classification based on eyes has many challenges. This gender classification is grouped into two classes, namely male and female. This study aims to build a gender classification model from eye image. The processes in this research include selecting the dataset, preprocessing the data, dividing the data which is divided into training data and test data, modeling, and evaluating the performance of the model. This study uses a public dataset, where the data contains a total of 2,681 images consisting of 1251 male eyes and 1430 female eyes. This study concludes that gender classification using eye image using the Inception-V3 method is better than the MobileNet method. This is obtained based on the accuracy value generated by the Inception-V3 method which is higher than the MobileNet-V2 method which obtains an accuracy of 91.82%.

Facial Age Estimation using Transfer Learning and Bayesian Optimization based on Gender Information

Abstract: Age estimation of unrestricted imaging circumstances has attracted an augmented recognition as it is appropriate in several real-world applications such as surveillance, face recognition, age synthesis, access control, and electronic customer relationship management. Current deep learning-based methods have displayed encouraging performance in age estimation field. Males and Females have a variable type of appearance aging pattern; this results in age differently. This fact leads to assuming that using gender information may improve the age estimator performance. We have proposed a novel model based on Gender Classification. A Convolutional Neural Network (CNN) is used to get Gender Information, then Bayesian Optimization is applied to this pre-trained CNN when fine-tuned for age estimation task. Bayesian Optimization reduces the classification error on the validation set for the pre-trained model. Extensive experiments are done to assess our proposed model on two data sets: FERET and FG-NET. The experiments’ result indicates that using a pre-trained CNN containing Gender Information with Bayesian Optimization outperforms the state of the arts on FERET and FG-NET data sets with a Mean Absolute Error (MAE) of 1.2 and 2.67 respectively.

Multibranch Convolutional Neural Network For Gender And Age Identification Using Multiclass Classification And FaceNet Model

Abstract: The human face provides a wealth of information regarding gender, age, ethnicity and emotions. Gender and age are considered as important biometrics and attributes for the identification process. However, the identification of gender and age is influenced by many dynamic factors that can change over time such as aging, hairstyles and expressions. The identification process have a problem in accuracy and the loss, several face recognition methodologies have been tried to overcome these dynamic factor problems, one of them is multibranch convolutional neural networks. The previous studies used these methods to deal with overfitting and backpropagation, but other supporting methods are still needed to increase the accuracy. The goal of this work is to optimize accuracy and mean absolute error (MAE), multiclass classification it's used to grouping data for age and facenet is used to solve problems related to face verification and overfitting, multibranch convolutional neural network (CNN) can be used to optimize backpropagation and reduce the error rate by adjusting weight based on the difference in output and the desired target.