School of Science and Technology 科技學院
Computing Programmes 電腦學系
| Programme | Bachelor of Computing with Honours in Internet Technology |
| Supervisor | Prof. Vanessa NG |
| Areas | Innovative Systems |
| Year of Completion | 2019 |
The project aims at facilitating teaching and learning in a smart learning environment by developing a roll-call system and an educational mobile application, enhancing the reliability of attendance list.
The traditional method of making roll-call by teacher is time-consuming and has an impact on the use of lesson time. Moreover, attendance taken by students may disturb students' learning, and the attendance list may not be credible. Besides, teachers' workloads will be increased for the administrative tasks to be done after class.
With the help of our roll-call system, attendance can be taken once student have arrived at classroom for generating an accurate attendance list. Also, students can use the mobile application to participate in class activities more. This can increase student class participation and create more chances of interactions between teachers and students.
In the name of facilitating a smart campus, the proposed educational application's objectives are sixfold:
Techniques and technologies used
Facial Recognition
Facial Recognition is a technology to find out human facial characteristics and shapes in an image or a video and recognize the identity of a person.
Figure 1: Facial Recognition Sample
Positioning
Positioning is used to calculate the position of an object in Space. Global Positioning System (GPS) is a famous positioning system to apply outdoor positioning.
Figure 2: Global Positioning System (GPS)
Wi-Fi positioning
Wi-Fi positioning system is a geolocation system to find out the locations of devices by connecting to nearby Wi-Fi hotspots and other wireless access points.
Figure 3: Indoor Wi-Fi Positioning
Roll-Call System
Create facial data set
During the registration process, students are required to take 3 available photos.
Roll-call by facial recognition
The surveillance camera can keep detecting faces, and the system will send the image to Face Recognition API for analyzing. Students' ID will then be returned, and the attendance record will be updated.
Roll-call by Wi-Fi positioning
The roll-call request will be sent if the mobile is connected to campus Wi-Fi, then the server will update the attendance record.
Smart Interacting System
Students can complete lab exercises prepared by teachers. Teachers can manage attendance, post exercises, and provide evaluations and comments in distributed exercises.
Wish Tree
Students can submit questions or opinions to teachers for further explanation or improvement.
Create facial data set
During the registration process, students are required to take 3 available photos. After that, the application will detect the number of facial features in the photos.
Figure 5: The procedure of creating facial data set
Roll-call by facial recognition
The surveillance camera can keep detecting faces, and after cropping the image, the system will send the image to Microsoft Cognitive Services Face API for analyzing. Students' ID will then be returned, and the attendance record in database will be updated.
Figure 6: The procedure of roll-call by facial recognition
Raspberry Pi is used as the surveillance camera. Raspberry Pi will process captured photos and send them to the server. In order to implement facial recognition technology, the combination between different extension modules and the Raspberry Pi computer is needed.
Figure 7: Raspberry Pi camera
When students first login, they need to take 3 valid photos by in-app function or select from photo library to create a facial data set. The application will authenticate the photos whether they belong to the same person.
Figure 8: Login page
Figure 9: Creating facial data set
After logging in, it will display students' locations and their subjects. With the help of facial recognition system, it processes the automatic roll-call and then releases confirmation notification. The attendance list is visible in the teacher's account which lights up the present students' icons after authentication.
Figure 10: Student’s Home page
Figure 11: Face detection in the classroom
Figure 12: Teacher’s attendance list page
When students press “subject”, related exercises will be listed, including the comment and the status of the exercises. Students can select the target exercise for completion. When finished, the status of the exercise will be updated as “done”.
Figure 13: Student’s Exercises List page
Figure 14: Exercises page
In user evaluations on our roll-call system, all respondents were satisfied with our roll-call system on its efficiency comparing to using attendance list. They believed our roll-call system could avoid proxy attendance and greatly reduced management workload.
Figure 15: Usefulness on Roll-call System
Some of the respondents felt that our interactive system could provide more chances for them to communicate with lecturers or students. Most of the respondents agreed with the facilitation of communication between teachers and students.
Figure 16: Usefulness on Interactive System
In the future, a Student-Side Model Training System will be designed to enhance the accuracy of face recognition.
Also, a Raspberry Pi camera setting plan should be made. It includes the consideration of the positions of multi-Raspberry Pi cameras to fit different sizes of lecture rooms.
Moreover, apart from the common short questions in the system, more detailed lab tests for professional subjects through interactive-multimedia, like 3D Biological Organs, can be included in the system.
