FusiformNet: Extracting Discriminative Facial Features on Different Levels
This addresses facial recognition accuracy for applications like security or identification, but it appears incremental as it builds on existing deep neural network approaches without a major paradigm shift.
The paper tackled the problem of extracting discriminative facial features for recognition by proposing FusiformNet, a framework that leverages both general and local facial differences, achieving a state-of-the-art accuracy of 96.67% on the Labeled Faces in the Wild benchmark without external data or specialized techniques.
Over the last several years, research on facial recognition based on Deep Neural Network has evolved with approaches like task-specific loss functions, image normalization and augmentation, network architectures, etc. However, there have been few approaches with attention to how human faces differ from person to person. Premising that inter-personal differences are found both generally and locally on the human face, I propose FusiformNet, a novel framework for feature extraction that leverages the nature of discriminative facial features. Tested on Image-Unrestricted setting of Labeled Faces in the Wild benchmark, this method achieved a state-of-the-art accuracy of 96.67% without labeled outside data, image augmentation, normalization, or special loss functions. Likewise, the method also performed on a par with previous state-of-the-arts when pre-trained on CASIA-WebFace dataset. Considering its ability to extract both general and local facial features, the utility of FusiformNet may not be limited to facial recognition but also extend to other DNN-based tasks.