A hybrid approach of neural networks for age and gender classification through decision fusion

A New Benchmark for Consumer Visual Tracking and Apparent Demographic Estimation from RGB and Thermal Images

Visual tracking and attribute estimation related to age or gender information of multiple person entities in a scene are mature research topics with the advent of deep learning techniques. However, when it comes to indoor images such as video sequences of retail consumers, data are not always adequate or accurate enough to essentially train effective models for consumer detection and tracking under various adverse factors. This in turn affects the quality of recognizing age or gender for those detected instances. In this work, we introduce two novel datasets: Consumers comprises 145 video sequences compliant to personal information regulations as far as facial images are concerned and BID is a set of cropped body images from each sequence that can be used for numerous computer vision tasks. We also propose an end-to-end framework which comprises CNNs as object detectors, LSTMs for motion forecasting of the tracklet association component in a sequence, along with a multi-attribute classification model for apparent demographic estimation of the detected outputs, aiming to capture useful metadata of consumer product preferences. Obtained results on tracking and age/gender prediction are promising with respect to reference systems while they indicate the proposed model's potential for practical consumer metadata extraction.

A space and time efficient convolutional neural network for age group estimation from facial images

Background

Age estimation has a wide range of applications, including security and surveillance, human-computer interaction, and biometrics. Facial aging is a stochastic process affected by various factors, such as lifestyle, habits, genetics, and the environment. Extracting age-related facial features to predict ages or age groups is a challenging problem that has attracted the attention of researchers in recent years. Various methods have been developed to solve the problem, including classification, regression-based methods, and soft computing approaches. Among these, the most successful results have been obtained by using neural network based artificial intelligence (AI) techniques such as convolutional neural networks (CNN). In particular, deep learning approaches have achieved improved accuracies by automatically extracting features from images of the human face. However, more improvements are still needed to achieve faster and more accurate results.

Methods

To address the aforementioned issues, this article proposes a space and time-efficient CNN method to extract distinct facial features from face images and classify them according to age group. The performance loss associated with using a small number of parameters to extract high-level features is compensated for by including a sufficient number of convolution layers. Additionally, we design and test suitable CNN structures that can handle smaller image sizes to assess the impact of size reduction on performance.

Results

To validate the proposed CNN method, we conducted experiments on the UTKFace and Facial-age datasets. The results demonstrated that the model outperformed recent studies in terms of classification accuracy and achieved an overall weighted F1-score of 87.84% for age-group classification problem.

A Deep Learning Method for Early Detection of Diabetic Foot Using Decision Fusion and Thermal Images

Diabetes mellitus (DM) is one of the major diseases that cause death worldwide and lead to complications of diabetic foot ulcers (DFU). Improper and late handling of a diabetic foot patient can result in an amputation of the patient’s foot. Early detection of DFU symptoms can be observed using thermal imaging with a computer-assisted classifier. Previous study of DFU detection using thermal image only achieved 97% of accuracy, and it has to be improved. This article proposes a novel framework for DFU classification based on thermal imaging using deep neural networks and decision fusion. Here, decision fusion combines the classification result from a parallel classifier. We used the convolutional neural network (CNN) model of ShuffleNet and MobileNetV2 as the baseline classifier. In developing the classifier model, firstly, the MobileNetV2 and ShuffleNet were trained using plantar thermogram datasets. Then, the classification results of those two models were fused using a novel decision fusion method to increase the accuracy rate. The proposed framework achieved 100% accuracy in classifying the DFU thermal images in binary classes of positive and negative cases. The accuracy of the proposed Decision Fusion (DF) was increased by about 3.4% from baseline ShuffleNet and MobileNetV2. Overall, the proposed framework outperformed in classifying the images compared with the state-of-the-art deep learning and the traditional machine-learning-based classifier.

Using Dynamic Pruned N-Gram Model for Identifying the Gender of the User

Organizations analyze customers’ personal data to understand and model their behavior. Identifying customers’ gender is a significant factor in analyzing markets that help plan the promotional campaigns, determine target customers and provide relevant offers. Several techniques were developed to analyze different types of data, including text, image, speech, and biometrics, to identify the gender of the user. The method of synthesis of the profile name differs from one customer to another. Using numerical substitutions of specific letters, known as Leet language, impedes the gender identification task. Moreover, using acronyms, misspellings, and adjacent names impose additional challenges. Towards this goal, this work uses the customers’ profile names associated with submitted reviews to recognize the customers’ gender. First, we create datasets of profile names extracted from the customers’ reviews. Secondly, we introduce a dynamic pruned n-gram model for identifying the gender of the user. It starts with data segmentation to handle adjacent parts, followed by data conversion and cleaning to fix the use of Leet language. Feature selection through a dynamic pruned n-gram model is the next step with the recurrent misspelling correction using fuzzy matching. We evaluate the proposed approach on the real data collected from active web resources. The obtained results demonstrate its validity and reliability.