Investigating bias in deep face analysis: The KANFace dataset and empirical study

Georgopoulos, Markos; Panagakis, Yannis; Pantić, Maja

doi:10.1016/j.imavis.2020.103954

Cited by 33 publications

(22 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• We also demonstrate that for age estimation, different facial parts have variable importance with "nose" being the least important region; • Our FP-Age achieves new state-of-the-art results on IMDB-Clean, Morph [5] and CACD [6]; • When trained on IMDB-Clean, our FP-Age also achieves state-of-the-art results on KANFace [7], FG-Net [8], Morph [5] and CACD [6] under cross-dataset evaluation.…”

mentioning

confidence: 67%

FP-Age: Leveraging Face Parsing Attention for Facial Age Estimation in the Wild

Lin¹,

Shen²,

Wang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Image-based age estimation aims to predict a person's age from facial images. It is used in a variety of real-world applications. Although end-to-end deep models have achieved impressive results for age estimation on benchmark datasets, their performance in-the-wild still leaves much room for improvement due to the challenges caused by large variations in head pose, facial expressions, and occlusions. To address this issue, we propose a simple yet effective method to explicitly incorporate facial semantics into age estimation, so that the model would learn to correctly focus on the most informative facial components from unaligned facial images regardless of head pose and nonrigid deformation. To this end, we design a face parsing-based network to learn semantic information at different scales and a novel face parsing attention module to leverage these semantic features for age estimation. To evaluate our method on in-the-wild data, we also introduce a new challenging large-scale benchmark called IMDB-Clean. This dataset is created by semi-automatically cleaning the noisy IMDB-WIKI dataset using a constrained clustering method. Through comprehensive experiment on IMDB-Clean and other benchmark datasets, under both intra-dataset and cross-dataset evaluation protocols, we show that our method consistently outperforms all existing age estimation methods and achieves a new state-of-the-art performance. To the best of our knowledge, our work presents the first attempt of leveraging face parsing attention to achieve semantic-aware age estimation, which may be inspiring to other high level facial analysis tasks.

show abstract

mentioning

confidence: 67%

FP-Age: Leveraging Face Parsing Attention for Facial Age Estimation in the Wild

Lin¹,

Shen²,

Wang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…More recent studies [26], [39], [65], [12], [36], [9], [34], [57], [29], [6] focused on jointly investigating the effects of user demographics on face recognition. These studies showed that the effects lead to an exponential face recognition error increase when facing the same biased race, gender, and age factors [34].…”

Section: A Estimating Bias In Face Recognitionmentioning

confidence: 99%

A Comprehensive Study on Face Recognition Biases Beyond Demographics

Terhörst

Kolf

Huber

et al. 2022

IEEE Trans. Technol. Soc.

View full text Add to dashboard Cite

Face recognition (FR) systems have a growing effect on critical decision-making processes. Recent works have shown that FR solutions show strong performance differences based on the user's demographics. However, to enable a trustworthy FR technology, it is essential to know the influence of an extended range of facial attributes on FR beyond demographics. Therefore, in this work, we analyse FR bias over a wide range of attributes. We investigate the influence of 47 attributes on the verification performance of two popular FR models. The experiments were performed on the publicly available MAAD-Face attribute database with over 120M high-quality attribute annotations. To prevent misleading statements about biased performances, we introduced control group based validity values to decide if unbalanced test data causes the performance differences. The results demonstrate that also many non-demographic attributes strongly affect the recognition performance, such as accessories, hair-styles and colors, face shapes, or facial anomalies. The observations of this work show the strong need for further advances in making FR system more robust, explainable, and fair. Moreover, our findings might help to a better understanding of how FR networks work, to enhance the robustness of these networks, and to develop more generalized bias-mitigating face recognition solutions.

show abstract

“…Georgopoulos et al [11] tested different deep learning based-face technology models such as facial recognition and gender prediction to find whether there is a bias in those systems or not. Moreover, the ten coding schemes method was used in [12] to analyze the diversity of subjects' faces in their proposed KANface dataset.…”

Section: Investigating Bias In Facial Data-based Technologymentioning

confidence: 99%

“…A method based on separating non-discriminative demographic features from the network embedding was proposed in [11] to mitigate bias resulted from unbalanced training data. However, it was useful only with bias toward age when applying face and gender recognition and it failed to remove gender bias for the age estimation and face recognition tasks.…”

Section: Mitigating Bias By Removing Demographic Featuresmentioning

confidence: 99%

A Study of Gender Bias in Face Presentation Attack and Its Mitigation

Alshareef¹,

Yuan²,

Atay³

2021

Future Internet

View full text Add to dashboard Cite

In biometric systems, the process of identifying or verifying people using facial data must be highly accurate to ensure a high level of security and credibility. Many researchers investigated the fairness of face recognition systems and reported demographic bias. However, there was not much study on face presentation attack detection technology (PAD) in terms of bias. This research sheds light on bias in face spoofing detection by implementing two phases. First, two CNN (convolutional neural network)-based presentation attack detection models, ResNet50 and VGG16 were used to evaluate the fairness of detecting imposer attacks on the basis of gender. In addition, different sizes of Spoof in the Wild (SiW) testing and training data were used in the first phase to study the effect of gender distribution on the models’ performance. Second, the debiasing variational autoencoder (DB-VAE) (Amini, A., et al., Uncovering and Mitigating Algorithmic Bias through Learned Latent Structure) was applied in combination with VGG16 to assess its ability to mitigate bias in presentation attack detection. Our experiments exposed minor gender bias in CNN-based presentation attack detection methods. In addition, it was proven that imbalance in training and testing data does not necessarily lead to gender bias in the model’s performance. Results proved that the DB-VAE approach (Amini, A., et al., Uncovering and Mitigating Algorithmic Bias through Learned Latent Structure) succeeded in mitigating bias in detecting spoof faces.

show abstract

Investigating bias in deep face analysis: The KANFace dataset and empirical study

Cited by 33 publications

References 16 publications

FP-Age: Leveraging Face Parsing Attention for Facial Age Estimation in the Wild

FP-Age: Leveraging Face Parsing Attention for Facial Age Estimation in the Wild

A Comprehensive Study on Face Recognition Biases Beyond Demographics

A Study of Gender Bias in Face Presentation Attack and Its Mitigation

Contact Info

Product

Resources

About