More Information Supervised Probabilistic Deep Face Embedding Learning
This work addresses the challenge of enhancing face embedding generalization for open-set recognition, representing an incremental improvement over existing margin-based methods.
The paper tackled the problem of improving generalization in open-set face recognition by analyzing margin-based softmax loss from a probability perspective and proposing two principles for designing new loss functions, resulting in a single model achieving over 99% accuracy on the MegaFace test.
Researches using margin based comparison loss demonstrate the effectiveness of penalizing the distance between face feature and their corresponding class centers. Despite their popularity and excellent performance, they do not explicitly encourage the generic embedding learning for an open set recognition problem. In this paper, we analyse margin based softmax loss in probability view. With this perspective, we propose two general principles: 1) monotonic decreasing and 2) margin probability penalty, for designing new margin loss functions. Unlike methods optimized with single comparison metric, we provide a new perspective to treat open set face recognition as a problem of information transmission. And the generalization capability for face embedding is gained with more clean information. An auto-encoder architecture called Linear-Auto-TS-Encoder(LATSE) is proposed to corroborate this finding. Extensive experiments on several benchmarks demonstrate that LATSE help face embedding to gain more generalization capability and it boosted the single model performance with open training dataset to more than $99\%$ on MegaFace test.