Log In Sign Up

NPT-Loss: A Metric Loss with Implicit Mining for Face Recognition

by   Syed Safwan Khalid, et al.

Face recognition (FR) using deep convolutional neural networks (DCNNs) has seen remarkable success in recent years. One key ingredient of DCNN-based FR is the appropriate design of a loss function that ensures discrimination between various identities. The state-of-the-art (SOTA) solutions utilise normalised Softmax loss with additive and/or multiplicative margins. Despite being popular, these Softmax+margin based losses are not theoretically motivated and the effectiveness of a margin is justified only intuitively. In this work, we utilise an alternative framework that offers a more direct mechanism of achieving discrimination among the features of various identities. We propose a novel loss that is equivalent to a triplet loss with proxies and an implicit mechanism of hard-negative mining. We give theoretical justification that minimising the proposed loss ensures a minimum separability between all identities. The proposed loss is simple to implement and does not require heavy hyper-parameter tuning as in the SOTA solutions. We give empirical evidence that despite its simplicity, the proposed loss consistently achieves SOTA performance in various benchmarks for both high-resolution and low-resolution FR tasks.


Support Vector Guided Softmax Loss for Face Recognition

Face recognition has witnessed significant progresses due to the advance...

Minimum Margin Loss for Deep Face Recognition

Face recognition has achieved great progress owing to the fast developme...

Mis-classified Vector Guided Softmax Loss for Face Recognition

Face recognition has witnessed significant progress due to the advances ...

Loss Function Search for Face Recognition

In face recognition, designing margin-based (e.g., angular, additive, ad...

A Decidability-Based Loss Function

Nowadays, deep learning is the standard approach for a wide range of pro...

SphereFace2: Binary Classification is All You Need for Deep Face Recognition

State-of-the-art deep face recognition methods are mostly trained with a...

Ring loss: Convex Feature Normalization for Face Recognition

We motivate and present Ring loss, a simple and elegant feature normaliz...