Learned Features are Better for Ethnicity Classification

Attique

et al. 2020

Sensors

Human face image analysis is an active research area within computer vision. In this paper we propose a framework for face image analysis, addressing three challenging problems of race, age, and gender recognition through face parsing. We manually labeled face images for training an end-to-end face parsing model through Deep Convolutional Neural Networks. The deep learning-based segmentation model parses a face image into seven dense classes. We use the probabilistic classification method and created probability maps for each face class. The probability maps are used as feature descriptors. We trained another Convolutional Neural Network model by extracting features from probability maps of the corresponding class for each demographic task (race, age, and gender). We perform extensive experiments on state-of-the-art datasets and obtained much better results as compared to previous results.

“…For classification, K-nearest neighbors (KNN) was used. Some more methods addressing race classification through holistic methods can be explored in [42][43][44].…”

Section: Race Classificationmentioning

confidence: 99%

Section: Race Classificationmentioning

confidence: 99%

See 1 more Smart Citation

A Multi-Task Framework for Facial Attributes Classification through End-to-End Face Parsing and Deep Convolutional Neural Networks

Attique

et al. 2020

Sensors

Artificial Neural Networks and Machine Learning – ICANN 2018

“…Inzamam et al [2] performed the classification by extracting features from a deep neural network followed by SVM classification on 10 datasets (13,394 images in total, including different variations of the FERET, CASPEAL, and Yale databases). Their classifier achieved success rates of 99.66%, 98.28%, and 99.05%, respectively, for the ethnic groups: African, Asian, and Caucasian.…”

Section: Recent Deep Learning Techniquesmentioning

confidence: 99%

DeepEthnic: Multi-label Ethnic Classification from Face Images

Huri

David

Netanyahu

2018

Ethnic group classification is a well-researched problem, which has been pursued mainly during the past two decades via traditional approaches of image processing and machine learning. In this paper, we propose a method of classifying an image face into an ethnic group by applying transfer learning from a previously trained classification network for large-scale data recognition. Our proposed method yields state-ofthe-art success rates of 99.02%, 99.76%, 99.2%, and 96.7%, respectively, for the four ethnic groups: African, Asian, Caucasian, and Indian.

“…Face providing several important information, including personal identity, gender [2], age [2][3][4][5], ethnicity [6,7] , skin tones and emotional expression [8][9][10][11]. Ethnicity identification is the key demographic attribute, and it plays an important role in many applications [12][13][14]. Thus, a human face must be detected and exploited in order to analyze the facial information.…”

Section: Introductionmentioning

confidence: 99%

A multi-color based features from facial images for automatic ethnicity identification model

Osman

Maarof

Rohani

et al. 2020

IJEECS

<span style="font-size: 9pt; font-family: 'Times New Roman', serif;">Ethnicity identification for demographic information has been studied for soft biometric analysis, and it is essential for human identification and verification. Ethnicity identification remains popular and receives attention in a recent year especially in automatic demographic information. Unfortunately, ethnicity identification technique using color-based feature mostly failed to determine the ethnicity classes accurately due to low properties of features in color-based. Thus, this paper purposely analyses the accuracy of the color-based ethnicity identification model from various color spaces. The proposed model involved several phases such as skin color feature extraction, feature selection, and classification. In the feature extraction process, a dynamic skin color detection is adapted to extract the skin color information from the face candidate. The multi-color feature was formed from the descriptive statistical model. Feature selection technique applied to reduce the feature space dimensionality. Finally, the proposed ethnicity identification was tested using several classification algorithms. From the experimental result, we achieved a better result in multi-color feature compared to individual color space model under Random Forest algorithm.</span>