Face Detection using Deep Learning: An Improved Faster RCNN Approach

TL;DR

This paper introduces an enhanced Faster R-CNN-based face detection method that integrates multiple strategies to achieve state-of-the-art performance on the FDDB benchmark.

Contribution

It improves the Faster R-CNN framework with techniques like feature concatenation and multi-scale training, setting new performance standards for face detection.

Findings

Achieved the best ROC curve performance on FDDB

Outperformed all previously published face detection methods

Demonstrated the effectiveness of combined strategies in deep learning models

Abstract

In this report, we present a new face detection scheme using deep learning and achieve the state-of-the-art detection performance on the well-known FDDB face detetion benchmark evaluation. In particular, we improve the state-of-the-art faster RCNN framework by combining a number of strategies, including feature concatenation, hard negative mining, multi-scale training, model pretraining, and proper calibration of key parameters. As a consequence, the proposed scheme obtained the state-of-the-art face detection performance, making it the best model in terms of ROC curves among all the published methods on the FDDB benchmark.