Facial Attractiveness Classification using Deep Learning

Anderson, Ricky; Gema, Aryo Pradipta; Suharjito, Suharjito; Isa, Sani Muhamad

doi:10.1109/inapr.2018.8627004

Cited by 8 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep convolutional neural networks (DCNNs) outperform handcrafted descriptors in terms of feature extraction and prediction. Some CNN framework-based models, such as VGG [17,18], ResNet [19], were used to represent the handcrafted features. A multi-task CNN called HMTNet was proposed in the study [20] that can predict the beauty score of a facial image besides race and gender utilizing SCUT-FBP and SCUT-FBP5500 datasets.…”

Section: Related Workmentioning

confidence: 99%

An Ensemble DCNNs-Based Regression Model for Automatic Facial Beauty Prediction and Analyzation

Saeed¹,

Abdulazeez²,

Ibrahim³

2023

View full text Add to dashboard Cite

One of the most effective social aspects of the human face is its attractiveness. Automatic facial beauty prediction (FBP) is an emerging research area that has gained much interest recently. However, identifying the significant facial traits and attributes that can contribute to the process of beauty attractiveness estimation is one of the main challenges in this research area. Furthermore, learning the beauty pattern from a relatively small, imbalanced dataset is another concern that needs to be addressed. This research proposes an ensemblebased regression model that integrates judgments made by three various DCNNs, each with a different structure representation. The proposed method efficiently predicts the beauty score by leveraging the strengths of each network as a complementary data source, and it draws attention to the most important beauty-related face features through the Gradientweighted Class Activation Mapping (Grad-CAM). The findings are promising, demonstrating the efficiency of fusing the decision of multiple predictors of the proposed ensemble DCNNs regression models that is significantly consistent with the ground truth of the employed datasets (SCUT-FBP, SCUT-FBP5500, and ME Beauty). Moreover, it can assist in comprehending the relationship between facial characteristics and the impression of attractiveness.

show abstract

Section: Related Workmentioning

confidence: 99%

An Ensemble DCNNs-Based Regression Model for Automatic Facial Beauty Prediction and Analyzation

Saeed¹,

Abdulazeez²,

Ibrahim³

2023

View full text Add to dashboard Cite

show abstract

“…FBP has received significant attention within computer vision as an emerging research area. Estimating the beauty level from a facial image could be treated as a classification [16][17][18], regression [19,20], or ranking [21,22] task. Two distinct categories of FBP exist.…”

Section: Related Workmentioning

confidence: 99%

Automatic Facial Aesthetic Prediction Based on Deep Learning with Loss Ensembles

Saeed,

Abdulazeez,

Ibrahim

2023

Applied Sciences

View full text Add to dashboard Cite

Deep data-driven methodologies have significantly enhanced the automatic facial beauty prediction (FBP), particularly convolutional neural networks (CNNs). However, despite its wide utilization in classification-based applications, the adoption of CNN in regression research is still constrained. In addition, biases in beauty scores assigned to facial images, such as preferences for specific, ethnicities, or age groups, present challenges to the effective generalization of models, which may not be appropriately addressed within conventional individual loss functions. Furthermore, regression problems commonly employ L2 loss to measure error rate, and this function is sensitive to outliers, making it difficult to generalize depending on the number of outliers in the training phase. Meanwhile, L1 loss is another regression-loss function that penalizes errors linearly and is less sensitive to outliers. The Log-cosh loss function is a flexible and robust loss function for regression problems. It provides a good compromise between the L1 and L2 loss functions. The Ensemble of multiple loss functions has been proven to improve the performance of deep-learning models in various tasks. In this work, we proposed to ensemble three regression-loss functions, namely L1, L2, and Log-cosh, and subsequently averaging them to create a new composite cost function. This strategy capitalizes on the unique traits of each loss function, constructing a unified framework that harmonizes outlier tolerance, precision, and adaptability. The proposed loss function’s effectiveness was demonstrated by incorporating it with three pretrained CNNs (AlexNet, VGG16-Net, and FIAC-Net) and evaluating it based on three FBP benchmarks (SCUT-FBP, SCUT-FBP5500, and MEBeauty). Integrating FIAC-Net with the proposed loss function yields remarkable outcomes across datasets due to its pretrained task of facial-attractiveness classification. The efficacy is evident in managing uncertain noise distributions, resulting in a strong correlation between machine- and human-rated aesthetic scores, along with low error rates.

show abstract

“…Traditional FBP algorithms [16] relied on facial landmarks and some combination of hand-engineered rules and shallow machine learning models. However since 2015, CNNs have dominated the FBP task [17,18,21,22,24,20,23,39] due to the wide availability of pretrained networks and increased access to public data. Gray et al [23] proposed a 4 layer CNN and were the first to discard facial landmarks.…”

Section: Related Workmentioning

confidence: 99%

“…Xu et al [21] proposed using models pretrained on other facial tasks as a starting point to address the lack of data for FBP. Anderson et al [24] benchmark a variety of CNN architectures on the CelebA dataset for binary attractiveness prediction. Both Fan et al [41] and Liu et al [25] propose replacing the regression output with a distribution prediction output and using a KL-Divergence loss rather than the standard mean squared error.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Photofeeler-D3: A Neural Network with Voter Modeling for Dating Photo Impression Prediction

Kalra¹,

Peterson²

2019

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

In just a few years, online dating has become the dominant way that young people meet to date, making the deceptively error-prone task of picking good dating profile photos vital to a generation's ability to form romantic connections. Until now, artificial intelligence approaches to Dating Photo Impression Prediction (DPIP) have been very inaccurate, unadaptable to real-world application, and have only taken into account a subject's physical attractiveness. To that effect, we propose Photofeeler-D3 -the first convolutional neural network as accurate as 10 human votes for how smart, trustworthy, and attractive the subject appears in highly variable dating photos. Our "attractive" output is also applicable to Facial Beauty Prediction (FBP), making Photofeeler-D3 state-of-the-art for both DPIP and FBP. We achieve this by leveraging Photofeeler's Dating Dataset (PDD) with over 1 million images and tens of millions of votes, our novel technique of voter modeling, and cutting-edge computer vision techniques.

show abstract

Facial Attractiveness Classification using Deep Learning

Cited by 8 publications

References 9 publications

An Ensemble DCNNs-Based Regression Model for Automatic Facial Beauty Prediction and Analyzation

An Ensemble DCNNs-Based Regression Model for Automatic Facial Beauty Prediction and Analyzation

Automatic Facial Aesthetic Prediction Based on Deep Learning with Loss Ensembles

Photofeeler-D3: A Neural Network with Voter Modeling for Dating Photo Impression Prediction

Contact Info

Product

Resources

About