Accurate Face Detection for High Performance

TL;DR

This paper enhances face detection performance, especially for tiny faces, by applying recent deep learning techniques and strategies to the RetinaNet framework, achieving state-of-the-art results on the WIDER FACE benchmark.

Contribution

It introduces a combination of recent tricks and modifications to RetinaNet, including IoU loss, two-step detection, data augmentation, max-out classification, and multi-scale testing, for improved face detection.

Findings

Achieves state-of-the-art performance on WIDER FACE dataset.

Effectively detects tiny faces with high accuracy.

Demonstrates the benefit of combining multiple recent techniques.

Abstract

Face detection has witnessed significant progress due to the advances of deep convolutional neural networks (CNNs). Its central issue in recent years is how to improve the detection performance of tiny faces. To this end, many recent works propose some specific strategies, redesign the architecture and introduce new loss functions for tiny object detection. In this report, we start from the popular one-stage RetinaNet approach and apply some recent tricks to obtain a high performance face detector. Specifically, we apply the Intersection over Union (IoU) loss function for regression, employ the two-step classification and regression for detection, revisit the data augmentation based on data-anchor-sampling for training, utilize the max-out operation for classification and use the multi-scale testing strategy for inference. As a consequence, the proposed face detection method achieves state-of-the-art performance on the most popular and challenging face detection benchmark WIDER FACE dataset.