ECML: An Ensemble Cascade Metric Learning Mechanism towards Face Verification

TL;DR

This paper introduces ECML, a hierarchical ensemble cascade metric learning framework for face verification that improves discriminative power and resists overfitting, outperforming existing methods.

Contribution

The paper proposes a novel ensemble cascade metric learning mechanism with hierarchical and ensemble strategies, including a robust Mahalanobis metric learning method with a closed-form solution.

Findings

EC-RMML outperforms state-of-the-art metric learning methods.

ECML effectively balances underfitting and overfitting.

The framework is applicable to various metric learning approaches.

Abstract

Face verification can be regarded as a 2-class fine-grained visual recognition problem. Enhancing the feature's discriminative power is one of the key problems to improve its performance. Metric learning technology is often applied to address this need, while achieving a good tradeoff between underfitting and overfitting plays the vital role in metric learning. Hence, we propose a novel ensemble cascade metric learning (ECML) mechanism. In particular, hierarchical metric learning is executed in the cascade way to alleviate underfitting. Meanwhile, at each learning level, the features are split into non-overlapping groups. Then, metric learning is executed among the feature groups in the ensemble manner to resist overfitting. Considering the feature distribution characteristics of faces, a robust Mahalanobis metric learning method (RMML) with closed-form solution is additionally proposed. It can avoid the computation failure issue on inverse matrix faced by some well-known metric learning approaches (e.g., KISSME). Embedding RMML into the proposed ECML mechanism, our metric learning paradigm (EC-RMML) can run in the one-pass learning manner. Experimental results demonstrate that EC-RMML is superior to state-of-the-art metric learning methods for face verification. And, the proposed ensemble cascade metric learning mechanism is also applicable to other metric learning approaches.