Data Uncertainty Learning in Face Recognition

Abstract

Modeling data uncertainty is important for noisy images, but seldom explored for face recognition. The pioneer work [35] considers uncertainty by modeling each face image embedding as a Gaussian distribution. It is quite effective. However, it uses fixed feature (mean of the Gaussian) from an existing model. It only estimates the variance and relies on an ad-hoc and costly metric. Thus, it is not easy to use. It is unclear how uncertainty affects feature learning.This work applies data uncertainty learning to face recognition, such that the feature (mean) and uncertainty (variance) are learnt simultaneously, for the first time. Two learning methods are proposed. They are easy to use and outperform existing deterministic methods as well as [35] on challenging unconstrained scenarios. We also provide insightful analysis on how incorporating uncertainty estimation helps reducing the adverse effects of noisy samples and affects the feature learning.

RCoNet: Deformable Mutual Information Maximization and High-order Uncertainty-aware Learning for Robust COVID-19 Detection

“…Aiming at combating the significant negative effects of uncertainty in deep neural networks, uncertainty learning has been getting lots of research attention, which facilitates the reliability assessment and solves risk-based decision-making problems [25], [26], [27]. In recent years, various frameworks have been proposed to characterize the uncertainty in the model parameters of deep neural networks, referred to as model uncertainty, due to the limited size of training data [28], [29], which can be reduced by collecting more training data [26], [30], [31].…”

“…Meanwhile, another kind of uncertainty in deep learning, referred to as data uncertainty, measures the noise inherent in given training data, and hence cannot be eliminated by having more training data [32]. To combat these two kinds of uncertainty, lots of works on various computer vision tasks, i.e., face recognition [25], semantic segmentation [33], object detection [34], person re-identification [35], etc., have introduced deep uncertainty learning to improve the robustness of deep learning model and interpretability of discriminant. For face recognition task in [26], an uncertainty-aware probabilistic face embedding (PFE) was proposed to represent face images as distributions by utilizing data uncertainty.…”

“…The presence of the image noise and label noise in CXR datasets may cause image latent representations generated by deep neural networks to be scattered in the entire feature space. To deal with this issue, [25], [26], [35] represent each image as a Gaussian distribution, that is defined by a mean (a standard feature vector) and a variance. However, the deep features of CXR samples we considered in this paper typically follow a complex, non-Gaussian distribution [41], [42], which cannot be fully captured by its first-order (mean) or second-order statistics (variance).…”

Face Image Quality Assessment: A Literature Survey

“…Shi and Jain [49], and based thereon Chang et al [43], proposed to compute an uncertainty vector that directly corresponds to the FR feature vector for a single face image. In other words, the two output vectors are representing the Gaussian variance and mean, respectively.…”

“…In addition, [49] explains how the uncertainty can be used to fuse embeddings for multiple images. Extending the concept of [49], [43] proposed two methods to learn both uncertainty (variance) and feature (mean) at the same time, without a separate uncertainty module. This means that the uncertainty can improve the overall training by reducing the influence of low quality images, which implies that the FR performance may improve even if the uncertainty isn't used after training, although it is noted that this kind of quality attention can reduce performance when only low quality cases are considered after training.…”