Deep learning based estimation of facial attributes on challenging mobile phone face datasets

The performance of face analysis and recognition systems depends on the quality of the acquired face data, which is influenced by numerous factors. Automatically assessing the quality of face data in terms of biometric utility can thus be useful to detect low-quality data and make decisions accordingly. This survey provides an overview of the face image quality assessment literature, which predominantly focuses on visible wavelength face image input. A trend towards deep learning based methods is observed, including notable conceptual differences among the recent approaches, such as the integration of quality assessment into face recognition models. Besides image selection, face image quality assessment can also be used in a variety of other application scenarios, which are discussed herein. Open issues and challenges are pointed out, i.a. highlighting the importance of comparability for algorithm evaluations, and the challenge for future work to create deep learning approaches that are interpretable in addition to providing accurate utility predictions.

show abstract

“…Reference Aspects Method(s) Datasets IJB-A, MS-Celeb-1M, CASIA-WebFace, LFW [4] DlDuat Same as [2], but i.a. with smartphone images.…”

Section: Factor-specific -Commonalitiesmentioning

confidence: 99%

Face Image Quality Assessment: A Literature Survey

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Rose and Bourlai evaluated DL and non-DL methods to determine three binary facial attributes in [45] and [48] (which is a continuation of [45] despite the publication date order): Whether the eyes are open or closed, whether there are glasses or not, and whether the face pose is mostly frontal or not. The two DL methods in both papers consisted of AlexNet [102] and GoogLeNet [103] (an incarnation of the Inception architecture), pretrained on ImageNet [122] data.…”

Section: DL Fqa Literaturementioning

confidence: 99%

“…A score-level fusion of a SVM and either AlexNet or GoogLeNet led to the best results in [45]. The evaluations in [48] employ a smartphone (iPhone 5S) dataset in addition to the non-smartphone data used in [45], the latter of which is only used for training. Of the non-DL methods, result values in [48] are only shown for the cubic kernel SVM approach, because the other methods performed worse.…”

Section: DL Fqa Literaturementioning

confidence: 99%

“…The evaluations in [48] employ a smartphone (iPhone 5S) dataset in addition to the non-smartphone data used in [45], the latter of which is only used for training. Of the non-DL methods, result values in [48] are only shown for the cubic kernel SVM approach, because the other methods performed worse. Whether the performance of the SVM or one of the two DL methods is better varies between the experiments of [48], which proposed to use the SVM trained on a combination of all used datasets (score-level fusion of SVM and one of the DL networks is not tested).…”

Section: DL Fqa Literaturementioning

confidence: 99%

See 1 more Smart Citation

Face Image Quality Assessment: A Literature Survey

Schlett,

Rathgeb,

Henniger

et al. 2020

Preprint

View full text Add to dashboard Cite

The performance of face analysis and recognition systems depends on the quality of the acquired face data, which is influenced by numerous factors. Automatically assessing the quality of face data in terms of biometric utility can thus be useful to filter out low quality data. This survey provides an overview of the face quality assessment literature in the framework of face biometrics, with a focus on face recognition based on visible wavelength face images as opposed to e.g. depth or infrared quality assessment. A trend towards deep learning based methods is observed, including notable conceptual differences among the recent approaches. Besides image selection, face image quality assessment can also be used in a variety of other application scenarios, which are discussed herein. Open issues and challenges are pointed out, i.a. highlighting the importance of comparability for algorithm evaluations, and the challenge for future work to create deep learning approaches that are interpretable in addition to providing accurate utility predictions.

show abstract

“…In some studies, these steps are replaced by applying a single deep neural network (DNN) architecture [2,3]. Facial recognition systems were applied in many computer systems, from real-time face tracking and recognition systems [4] and identification of people in personal smartphone galleries [5] to large-scale classification and security systems [6]. Chang et al [7] proposed a facial recognition algorithm based on a support vector machine (SVM) combined with Visual Geometry Group (VGG) network model for extracting facial features, which not only accurately extracts face features, but also reduces feature dimensions and avoids irrelevant features in the calculation.…”

Section: Face Recognitionmentioning

confidence: 99%

A Method for Human Facial Image Annotation on Low Power Consumption Autonomous Devices

Hachaj

2020

Sensors

View full text Add to dashboard Cite

This paper proposes a classifier designed for human facial feature annotation, which is capable of running on relatively cheap, low power consumption autonomous microcomputer systems. An autonomous system is one that depends only on locally available hardware and software—for example, it does not use remote services available through the Internet. The proposed solution, which consists of a Histogram of Oriented Gradients (HOG) face detector and a set of neural networks, has comparable average accuracy and average true positive and true negative ratio to state-of-the-art deep neural network (DNN) architectures. However, contrary to DNNs, it is possible to easily implement the proposed method in a microcomputer with very limited RAM memory and without the use of additional coprocessors. The proposed method was trained and evaluated on a large 200,000 image face data set and compared with results obtained by other researchers. Further evaluation proves that it is possible to perform facial image attribute classification using the proposed algorithm on incoming video data captured by an RGB camera sensor of the microcomputer. The obtained results can be easily reproduced, as both the data set and source code can be downloaded. Developing and evaluating the proposed facial image annotation algorithm and its implementation, which is easily portable between various hardware and operating systems (virtually the same code works both on high-end PCs and microcomputers using the Windows and Linux platforms) and which is dedicated for low power consumption devices without coprocessors, is the main and novel contribution of this research.

show abstract

Deep learning based estimation of facial attributes on challenging mobile phone face datasets

Cited by 10 publications

References 30 publications

Face Image Quality Assessment: A Literature Survey

Face Image Quality Assessment: A Literature Survey

Face Image Quality Assessment: A Literature Survey

A Method for Human Facial Image Annotation on Low Power Consumption Autonomous Devices

Contact Info

Product

Resources

About