Generic Image Classification Approaches Excel on Face Recognition

TL;DR

This work demonstrates that standard image classification pipelines outperform specialized face recognition methods on multiple benchmarks, suggesting face recognition can leverage generic image classification techniques.

Contribution

The paper shows that generic image classification approaches, including simple dictionary methods, outperform state-of-the-art face recognition techniques on benchmark datasets.

Findings

Standard classification pipeline outperforms specialized face recognition methods.

Random patch dictionaries are effective for face recognition.

Learning multiple feature-specific dictionaries improves accuracy.

Abstract

The main finding of this work is that the standard image classification pipeline, which consists of dictionary learning, feature encoding, spatial pyramid pooling and linear classification, outperforms all state-of-the-art face recognition methods on the tested benchmark datasets (we have tested on AR, Extended Yale B, the challenging FERET, and LFW-a datasets). This surprising and prominent result suggests that those advances in generic image classification can be directly applied to improve face recognition systems. In other words, face recognition may not need to be viewed as a separate object classification problem. While recently a large body of residual based face recognition methods focus on developing complex dictionary learning algorithms, in this work we show that a dictionary of randomly extracted patches (even from non-face images) can achieve very promising results using the image classification pipeline. That means, the choice of dictionary learning methods may not be important. Instead, we find that learning multiple dictionaries using different low-level image features often improve the final classification accuracy. Our proposed face recognition approach offers the best reported results on the widely-used face recognition benchmark datasets. In particular, on the challenging FERET and LFW-a datasets, we improve the best reported accuracies in the literature by about 20% and 30% respectively.