The problem with distance learning is that it is not possible to supervise the student in the usual ways. In addition, the communication channel between teacher and student is narrowed, making it more difficult for student feedback to reach the teacher. In my solution I would use one of my hobby projects that I have been developing for 3 years.
This system is currently designed for an automotive environment, but it can also be used to bridge the communication gap between students and teachers. Video of the system: https://bit.ly/39GM2uX One of the problems my system can solve is to keep the students' attention during lectures.
My system is able to
It can measure the level of concentration by monitoring eye opening and eye movement frequency to determine how attentive the student is to the lecture.
It can detect yawning and signs of fatigue, etc.
Identify emotions, so it can be used to record students' feelings about a lecturer or lesson.
If we look at how an influencer communicates, we find that they are constantly talking. There is hardly a moment's pause in communication.
In contrast, a traditional teacher's communication is not continuous, sometimes taking longer pauses to think about what he or she is saying. Then the students' attention is immediately lost and they move on to something else. My system can help to detect this, giving you the opportunity to bring the students' attention back to the lecture. * By observing students (measuring their activity), the system can also help you find distracted or hyperactive children, so they can get extra help. The other problem that can be solved with my system is the examination. It can be used for high stakes exams to secure the exam and detect cheating.
The system is able to:
Facial recognition, so it can determine whether the right person is sitting in front of the machine or another person assisting the examinee.
It can detect mouth movement, so it can tell if the examinee is communicating with an assistant.
It can track gaze so it can detect if the student is using an aid and not looking at the monitor.
Usually, such exams use an online proctor who supervises (monitors) the candidate during the exam.With my system, a semi-automatic solution can be implemented, where the system alerts the proctor when suspicious events are detected. This way, a proctor can monitor more than one examinee at a time. This makes it possible to monitor a large number of candidates with fewer proctors, which is more cost-effective. My system may also be able to determine whether the stream from the student/examiner is live or pre-recorded. This is less of a problem for traditional education, but can be a serious problem for exams. For example, in the cases below, a simple trick is used to fool the teacher/examiner:
https://bit.ly/3dLB8qO
https://bit.ly/2V3A9tL
The system can also fit an upper body skeleton based on the image from the camera. By asking the student to perform certain actions (e.g., raise one hand), the fact that they do so can confirm that a live stream is being seen by the observer on the other side.