Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning
Wearing masks in indoor and outdoor public places has been mandatory in a number of countries during the COVID-19 pandemic. Correctly wearing a face mask can reduce the transmission of the virus through respiratory droplets. In this paper, a novel two-step deep learning (DL) method based on our exte...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-12-01
|
| Series: | Computers |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-431X/13/1/7 |
| _version_ | 1797344348318203904 |
|---|---|
| author | Yuan Zhang Meysam Effati Aaron Hao Tan Goldie Nejat |
| author_facet | Yuan Zhang Meysam Effati Aaron Hao Tan Goldie Nejat |
| author_sort | Yuan Zhang |
| collection | DOAJ |
| description | Wearing masks in indoor and outdoor public places has been mandatory in a number of countries during the COVID-19 pandemic. Correctly wearing a face mask can reduce the transmission of the virus through respiratory droplets. In this paper, a novel two-step deep learning (DL) method based on our extended ResNet-50 is presented. It can detect and classify whether face masks are missing, are worn correctly or incorrectly, or the face is covered by other means (e.g., a hand or hair). Our DL method utilizes transfer learning with pretrained ResNet-50 weights to reduce training time and increase detection accuracy. Training and validation are achieved using the MaskedFace-Net, MAsked FAces (MAFA), and CelebA datasets. The trained model has been incorporated onto a socially assistive robot for robust and autonomous detection by a robot using lower-resolution images from the onboard camera. The results show a classification accuracy of 84.13% for the classification of no mask, correctly masked, and incorrectly masked faces in various real-world poses and occlusion scenarios using the robot. |
| first_indexed | 2024-03-08T11:01:03Z |
| format | Article |
| id | doaj.art-4bf5f95c78104faeb9325e39d74aa850 |
| institution | Directory Open Access Journal |
| issn | 2073-431X |
| language | English |
| last_indexed | 2024-03-08T11:01:03Z |
| publishDate | 2023-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Computers |
| spelling | doaj.art-4bf5f95c78104faeb9325e39d74aa8502024-01-26T15:52:23ZengMDPI AGComputers2073-431X2023-12-01131710.3390/computers13010007Robust Face Mask Detection by a Socially Assistive Robot Using Deep LearningYuan Zhang0Meysam Effati1Aaron Hao Tan2Goldie Nejat3Autonomous Systems and Biomechatronics Laboratory (ASBLab), Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON M5S 3G8, CanadaAutonomous Systems and Biomechatronics Laboratory (ASBLab), Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON M5S 3G8, CanadaAutonomous Systems and Biomechatronics Laboratory (ASBLab), Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON M5S 3G8, CanadaAutonomous Systems and Biomechatronics Laboratory (ASBLab), Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON M5S 3G8, CanadaWearing masks in indoor and outdoor public places has been mandatory in a number of countries during the COVID-19 pandemic. Correctly wearing a face mask can reduce the transmission of the virus through respiratory droplets. In this paper, a novel two-step deep learning (DL) method based on our extended ResNet-50 is presented. It can detect and classify whether face masks are missing, are worn correctly or incorrectly, or the face is covered by other means (e.g., a hand or hair). Our DL method utilizes transfer learning with pretrained ResNet-50 weights to reduce training time and increase detection accuracy. Training and validation are achieved using the MaskedFace-Net, MAsked FAces (MAFA), and CelebA datasets. The trained model has been incorporated onto a socially assistive robot for robust and autonomous detection by a robot using lower-resolution images from the onboard camera. The results show a classification accuracy of 84.13% for the classification of no mask, correctly masked, and incorrectly masked faces in various real-world poses and occlusion scenarios using the robot.https://www.mdpi.com/2073-431X/13/1/7socially assistive robotsautonomous face mask detectionhuman–robot interactionsCOVID-19 pandemicextended ResNet-50 |
| spellingShingle | Yuan Zhang Meysam Effati Aaron Hao Tan Goldie Nejat Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning Computers socially assistive robots autonomous face mask detection human–robot interactions COVID-19 pandemic extended ResNet-50 |
| title | Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning |
| title_full | Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning |
| title_fullStr | Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning |
| title_full_unstemmed | Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning |
| title_short | Robust Face Mask Detection by a Socially Assistive Robot Using Deep Learning |
| title_sort | robust face mask detection by a socially assistive robot using deep learning |
| topic | socially assistive robots autonomous face mask detection human–robot interactions COVID-19 pandemic extended ResNet-50 |
| url | https://www.mdpi.com/2073-431X/13/1/7 |
| work_keys_str_mv | AT yuanzhang robustfacemaskdetectionbyasociallyassistiverobotusingdeeplearning AT meysameffati robustfacemaskdetectionbyasociallyassistiverobotusingdeeplearning AT aaronhaotan robustfacemaskdetectionbyasociallyassistiverobotusingdeeplearning AT goldienejat robustfacemaskdetectionbyasociallyassistiverobotusingdeeplearning |