Learning to Drop Expensive Layers for Fast Face Recognition

Recent years have seen many advances based on Deep Convolutional Neural Networks (DCNNs) in the tasks of face recognition, most of which are developed to pursue high recognition accuracy. In this paper, we propose a novel Fast FAce Recognizer (Fast-FAR), learning to improve the speed of DCNN-based f...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Junhui Li, Wei Jia, Yan Hu, Shouqing Li, Xiaoguang Tu
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Fast face recognition reinforcement learning deep convolutional neural networks Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/9d563da02e5b484391cd08001d635784
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:9d563da02e5b484391cd08001d635784
record_format	dspace
spelling	oai:doaj.org-article:9d563da02e5b484391cd08001d6357842021-11-19T00:06:57ZLearning to Drop Expensive Layers for Fast Face Recognition2169-353610.1109/ACCESS.2021.3106483https://doaj.org/article/9d563da02e5b484391cd08001d6357842021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9520823/https://doaj.org/toc/2169-3536Recent years have seen many advances based on Deep Convolutional Neural Networks (DCNNs) in the tasks of face recognition, most of which are developed to pursue high recognition accuracy. In this paper, we propose a novel Fast FAce Recognizer (Fast-FAR), learning to improve the speed of DCNN-based face recognition model without sacrificing recognition accuracy. Our fundamental insight is that the computation increases exponentially with the depth of a network, the easily identifiable face images can be accurately recognized by the cheap features (pixel values at shallow layers), while the challenging samples that exhibit low quality, large pose variations or occlusions need to be processed by the expensive deep layers. The major contribution of this paper is the Reinforcement Learning Agent (RLA), which is proposed to learn a decision policy determined by a reward function. The policy adaptively decides whether the recognition should be performed at an early layer with a high recognition confidence, or proceeding to the subsequent layers, thus significantly reducing feed-forward cost for the easy faces. According to the extensive experiments on the popular face recognition benchmarks, Fast-FAR reduces the inference time by 14.22%, 20.61%, and 7.84% on the dataset LFW, AgeDB-30 and CFP-FP, respectively.Junhui LiWei JiaYan HuShouqing LiXiaoguang TuIEEEarticleFast face recognitionreinforcement learningdeep convolutional neural networksElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 117880-117886 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Fast face recognition reinforcement learning deep convolutional neural networks Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Fast face recognition reinforcement learning deep convolutional neural networks Electrical engineering. Electronics. Nuclear engineering TK1-9971 Junhui Li Wei Jia Yan Hu Shouqing Li Xiaoguang Tu Learning to Drop Expensive Layers for Fast Face Recognition
description	Recent years have seen many advances based on Deep Convolutional Neural Networks (DCNNs) in the tasks of face recognition, most of which are developed to pursue high recognition accuracy. In this paper, we propose a novel Fast FAce Recognizer (Fast-FAR), learning to improve the speed of DCNN-based face recognition model without sacrificing recognition accuracy. Our fundamental insight is that the computation increases exponentially with the depth of a network, the easily identifiable face images can be accurately recognized by the cheap features (pixel values at shallow layers), while the challenging samples that exhibit low quality, large pose variations or occlusions need to be processed by the expensive deep layers. The major contribution of this paper is the Reinforcement Learning Agent (RLA), which is proposed to learn a decision policy determined by a reward function. The policy adaptively decides whether the recognition should be performed at an early layer with a high recognition confidence, or proceeding to the subsequent layers, thus significantly reducing feed-forward cost for the easy faces. According to the extensive experiments on the popular face recognition benchmarks, Fast-FAR reduces the inference time by 14.22%, 20.61%, and 7.84% on the dataset LFW, AgeDB-30 and CFP-FP, respectively.
format	article
author	Junhui Li Wei Jia Yan Hu Shouqing Li Xiaoguang Tu
author_facet	Junhui Li Wei Jia Yan Hu Shouqing Li Xiaoguang Tu
author_sort	Junhui Li
title	Learning to Drop Expensive Layers for Fast Face Recognition
title_short	Learning to Drop Expensive Layers for Fast Face Recognition
title_full	Learning to Drop Expensive Layers for Fast Face Recognition
title_fullStr	Learning to Drop Expensive Layers for Fast Face Recognition
title_full_unstemmed	Learning to Drop Expensive Layers for Fast Face Recognition
title_sort	learning to drop expensive layers for fast face recognition
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/9d563da02e5b484391cd08001d635784
work_keys_str_mv	AT junhuili learningtodropexpensivelayersforfastfacerecognition AT weijia learningtodropexpensivelayersforfastfacerecognition AT yanhu learningtodropexpensivelayersforfastfacerecognition AT shouqingli learningtodropexpensivelayersforfastfacerecognition AT xiaoguangtu learningtodropexpensivelayersforfastfacerecognition
_version_	1718420625624662016

Learning to Drop Expensive Layers for Fast Face Recognition

Ejemplares similares