Progressive boosting for class imbalance and its application to face re-identification

Abstract

In practice, pattern recognition applications often suffer from imbalanced data distributions between classes, which may vary during operations w.r.t. the design data. For instance, in many video surveillance applications, e.g., face re-identification, the face individuals must be recognized over a distributed network of video cameras. An important challenge in such applications is class imbalance since the number of faces captured from an individual of interest is greatly outnumbered by those of others. Two-class classification systems designed using imbalanced data tend to recognize the majority (negative) class better, while the class of interest (positive class) often has the smaller number of samples. Several data-level techniques have been proposed to alleviate this issue, where classifier ensembles are designed with balanced data subsets by up-sampling positive samples or under-sampling negative samples. However, some informative samples may be neglected by random under-sampling and adding synthetic positive samples through up-sampling adds to training complexity. In this paper, a new ensemble learning algorithm called Progressive Boosting (PBoost) is proposed that progressively inserts uncorrelated groups of samples into a Boosting procedure to avoid loosing information while generating a diverse pool of classifiers. In many real-world recognition problems, the samples may be regrouped using some application-based contextual information. For example, in face re-identification applications, facial regions of a same person appearing in a camera field of view may be regrouped based on their trajectories found by face tracker. From one iteration to the next, the PBoost algorithm accumulates these uncorrelated groups of samples into a set that grows gradually in size and imbalance. Base classifiers are trained on samples selected from this set and validated on the whole set. Consequently, PBoost is more robust when the operational data may have unknown and variable levels of skew. In addition, the computation complexity of PBoost is lower than Boosting ensembles in literature that use under-sampling for learning from imbalanced data because not all of the base classifiers are validated on all negative samples. The new loss factor used in PBoost avoids biasing performance towards the negative class. Using this loss factor, the weight update of samples and classifier contribution in final predictions are set according to the ability of classifiers to recognize both classes. The proposed approach was validated and compared using synthetic data and videos from the Faces In Action, and COX dataset that emulate face re-identification applications. Results show that PBoost outperforms state of the art techniques in terms of both accuracy and complexity over different levels of imbalance and overlap between classes.

Hybrid approach redefinition with progressive boosting for class imbalance problem

“…Weighted matrices for positive samples denoted by ,+ and Negative matrices for negative samples denoted by ,− . Weighted matrices for positive samples and negative samples can be calculated using (1) to (2) [15].…”

“…It is intended that information is not lost and can produce a collection of various classifications. Based on this, the PBoost method is expected to improve diversity data [15]. This research will combine the Hybrid Approach Redefinition by replacing the use of SMOTE Boost by using Progressive Boosting to get better data diversity, a small number of classifiers, and better performance.…”

Facial Action Units for Training Convolutional Neural Networks

“…However, it is often difficult to have not only sufficient but also class-balanced data. In particular, the data shortage and the class-imbalance problems are severe for most domain-specific applications such as bankruptcy prediction [18], action recognition [19], face re-identification [20], cancer detection [21], fraud detection [22], chemical and biomedical engineering [23], financial management [24], and diseases diagnosis [25]. In fact, the class-imbalance is a commonly encountered problem for classical pattern recognition problems and the most popular approach to solve the imbalance problem relies on the sampling methods.…”

Imbalanced biomedical data classification using self-adaptive multilayer ELM combined with dynamic GAN

“…Additionally, the training set sometimes contains high-dimensional feature and small samples. These factors further result in a lower classification accuracy of abnormal class and incorrect diagnosis result [ 4 ]. Therefore, establishing an effective classification model is an urgently necessary task for limited and imbalanced biomedical dataset.…”