Metric Classification Network in Actual Face Recognition Scene
This addresses a practical issue in real-world face recognition applications by eliminating the need for threshold adjustments, though it is incremental as it builds on existing models.
The paper tackles the problem of needing to adjust thresholds for face verification across different test sets by training a validation classifier to normalize the decision threshold, achieving relative error reductions of 25.37% on LFW and 26.60% on YTF compared to traditional methods.
In order to make facial features more discriminative, some new models have recently been proposed. However, almost all of these models use the traditional face verification method, where the cosine operation is performed using the features of the bottleneck layer output. However, each of these models needs to change a threshold each time it is operated on a different test set. This is very inappropriate for application in real-world scenarios. In this paper, we train a validation classifier to normalize the decision threshold, which means that the result can be obtained directly without replacing the threshold. We refer to our model as validation classifier, which achieves best result on the structure consisting of one convolution layer and six fully connected layers. To test our approach, we conduct extensive experiments on Labeled Face in the Wild (LFW) and Youtube Faces (YTF), and the relative error reduction is 25.37% and 26.60% than traditional method respectively. These experiments confirm the effectiveness of validation classifier on face recognition task.