Feature Agglomeration Networks for Single Stage Face Detection

TL;DR

This paper introduces Feature Agglomeration Networks (FANet), a single-stage face detection framework that leverages multi-scale features for improved accuracy and efficiency, achieving state-of-the-art results on multiple benchmarks.

Contribution

The paper proposes a novel hierarchical feature aggregation framework and a new loss function for real-time, high-accuracy face detection in unconstrained environments.

Findings

Achieved state-of-the-art performance on PASCAL face, FDDB, and WIDER FACE datasets.

Runs in real-time on VGA-resolution images on GPU.

Effectively handles faces at various scales with hierarchical feature aggregation.

Abstract

Recent years have witnessed promising results of face detection using deep learning. Despite making remarkable progresses, face detection in the wild remains an open research challenge especially when detecting faces at vastly different scales and characteristics. In this paper, we propose a novel simple yet effective framework of "Feature Agglomeration Networks" (FANet) to build a new single stage face detector, which not only achieves state-of-the-art performance but also runs efficiently. As inspired by Feature Pyramid Networks (FPN), the key idea of our framework is to exploit inherent multi-scale features of a single convolutional neural network by aggregating higher-level semantic feature maps of different scales as contextual cues to augment lower-level feature maps via a hierarchical agglomeration manner at marginal extra computation cost. We further propose a Hierarchical Loss to effectively train the FANet model. We evaluate the proposed FANet detector on several public face detection benchmarks, including PASCAL face, FDDB and WIDER FACE datasets and achieved state-of-the-art results. Our detector can run in real time for VGA-resolution images on GPU.