# MSFD:Multi-Scale Receptive Field Face Detector

## TL;DR

MSFD is a real-time face detector leveraging multi-scale receptive fields, hierarchical context, and an anchor strategy, achieving high accuracy across various datasets with efficient inference speed.

## Contribution

The paper introduces MSFD, a novel face detection method that combines multi-scale receptive fields, a new anchor assignment strategy, and focal loss for improved accuracy and speed.

## Key findings

- Achieves superior detection performance on FDDB, Pascal-Faces, and WIDER FACE datasets.
- Runs at 31 FPS on GPU for VGA images.
- Effectively detects faces of various scales, including small and rotated faces.

## Abstract

We aim to study the multi-scale receptive fields of a single convolutional neural network to detect faces of varied scales. This paper presents our Multi-Scale Receptive Field Face Detector (MSFD), which has superior performance on detecting faces at different scales and enjoys real-time inference speed. MSFD agglomerates context and texture by hierarchical structure. More additional information and rich receptive field bring significant improvement but generate marginal time consumption. We simultaneously propose an anchor assignment strategy which can cover faces with a wide range of scales to improve the recall rate of small faces and rotated faces. To reduce the false positive rate, we train our detector with focal loss which keeps the easy samples from overwhelming. As a result, MSFD reaches superior results on the FDDB, Pascal-Faces and WIDER FACE datasets, and can run at 31 FPS on GPU for VGA-resolution images.