Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning
Estimating the number of people in highly clustered crowd scenes is an extremely challenging task on account of serious occlusion and non-uniformity distribution in one crowd image. Traditional works on crowd counting take advantage of different CNN like networks to regress crowd density map, and fu...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-03-01
|
Series: | Sensors |
Subjects: | |
Online Access: | http://www.mdpi.com/1424-8220/19/6/1346 |
id |
doaj-aa4c0266b33d4d8c82b4172f1234cda0 |
---|---|
record_format |
Article |
spelling |
doaj-aa4c0266b33d4d8c82b4172f1234cda02020-11-25T03:26:22ZengMDPI AGSensors1424-82202019-03-01196134610.3390/s19061346s19061346Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep LearningMinglei Tong0Lyuyuan Fan1Hao Nan2Yan Zhao3School of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, ChinaSchool of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, ChinaSchool of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, ChinaSchool of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200090, ChinaEstimating the number of people in highly clustered crowd scenes is an extremely challenging task on account of serious occlusion and non-uniformity distribution in one crowd image. Traditional works on crowd counting take advantage of different CNN like networks to regress crowd density map, and further predict the count. In contrast, we investigate a simple but valid deep learning model that concentrates on accurately predicting the density map and simultaneously training a density level classifier to relax parameters of the network to prevent dangerous stampede with a smart camera. First, a combination of atrous and fractional stride convolutional neural network (CAFN) is proposed to deliver larger receptive fields and reduce the loss of details during down-sampling by using dilated kernels. Second, the expanded architecture is offered to not only precisely regress the density map, but also classify the density level of the crowd in the meantime (MTCAFN, multiple tasks CAFN for both regression and classification). Third, experimental results demonstrated on four datasets (Shanghai Tech A (MAE = 88.1) and B (MAE = 18.8), WorldExpo’10(average MAE = 8.2), NS UCF_CC_50(MAE = 303.2) prove our proposed method can deliver effective performance.http://www.mdpi.com/1424-8220/19/6/1346crowd countingmultiple task learningsmart camerafractional stride network |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Minglei Tong Lyuyuan Fan Hao Nan Yan Zhao |
spellingShingle |
Minglei Tong Lyuyuan Fan Hao Nan Yan Zhao Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning Sensors crowd counting multiple task learning smart camera fractional stride network |
author_facet |
Minglei Tong Lyuyuan Fan Hao Nan Yan Zhao |
author_sort |
Minglei Tong |
title |
Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning |
title_short |
Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning |
title_full |
Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning |
title_fullStr |
Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning |
title_full_unstemmed |
Smart Camera Aware Crowd Counting via Multiple Task Fractional Stride Deep Learning |
title_sort |
smart camera aware crowd counting via multiple task fractional stride deep learning |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2019-03-01 |
description |
Estimating the number of people in highly clustered crowd scenes is an extremely challenging task on account of serious occlusion and non-uniformity distribution in one crowd image. Traditional works on crowd counting take advantage of different CNN like networks to regress crowd density map, and further predict the count. In contrast, we investigate a simple but valid deep learning model that concentrates on accurately predicting the density map and simultaneously training a density level classifier to relax parameters of the network to prevent dangerous stampede with a smart camera. First, a combination of atrous and fractional stride convolutional neural network (CAFN) is proposed to deliver larger receptive fields and reduce the loss of details during down-sampling by using dilated kernels. Second, the expanded architecture is offered to not only precisely regress the density map, but also classify the density level of the crowd in the meantime (MTCAFN, multiple tasks CAFN for both regression and classification). Third, experimental results demonstrated on four datasets (Shanghai Tech A (MAE = 88.1) and B (MAE = 18.8), WorldExpo’10(average MAE = 8.2), NS UCF_CC_50(MAE = 303.2) prove our proposed method can deliver effective performance. |
topic |
crowd counting multiple task learning smart camera fractional stride network |
url |
http://www.mdpi.com/1424-8220/19/6/1346 |
work_keys_str_mv |
AT mingleitong smartcameraawarecrowdcountingviamultipletaskfractionalstridedeeplearning AT lyuyuanfan smartcameraawarecrowdcountingviamultipletaskfractionalstridedeeplearning AT haonan smartcameraawarecrowdcountingviamultipletaskfractionalstridedeeplearning AT yanzhao smartcameraawarecrowdcountingviamultipletaskfractionalstridedeeplearning |
_version_ |
1724593156596432896 |