Adversarial Focal Loss: Asking Your Discriminator for Hard Examples

TL;DR

This paper introduces Adversarial Focal Loss, a novel method that uses an adversarial network to identify hard examples for improved keypoint detection, extending Focal Loss beyond classification tasks.

Contribution

It proposes AFL, a versatile adaptation of Focal Loss that leverages an adversarial network to dynamically prioritize hard examples in keypoint detection.

Findings

AFL improves keypoint detection performance.

AFL effectively re-weights examples based on difficulty.

AFL generalizes Focal Loss to non-classification tasks.

Abstract

Focal Loss has reached incredible popularity as it uses a simple technique to identify and utilize hard examples to achieve better performance on classification. However, this method does not easily generalize outside of classification tasks, such as in keypoint detection. In this paper, we propose a novel adaptation of Focal Loss for keypoint detection tasks, called Adversarial Focal Loss (AFL). AFL not only is semantically analogous to Focal loss, but also works as a plug-and-chug upgrade for arbitrary loss functions. While Focal Loss requires output from a classifier, AFL leverages a separate adversarial network to produce a difficulty score for each input. This difficulty score can then be used to dynamically prioritize learning on hard examples, even in absence of a classifier. In this work, we show AFL's effectiveness in enhancing existing methods in keypoint detection and verify its capability to re-weigh examples based on difficulty.