Core demands
In recent years, children's teaching quality and campus safety have become a hot issue of increasing concern in the community. In addition, protecting every child is also an important part of safety education and management in primary and secondary schools. Therefore, at present, there are two problems that need to be solved urgently on campus:
1. At present, the teaching management mode is backward, the teaching quality evaluation relies on human perception, lacks scientific statistical means, and the evaluation tracking cycle is too long to adjust the teaching plan in time;
2. The campus mainly relies on the way of alternate patrolling between the gate guards for screening and the security personnel at fixed time, and there are problems such as the mixing of non campus personnel and the security loopholes of personnel leaving the post. This will directly lead to high randomness, high labor costs, low efficiency and poor security of campus security incidents.
Solution
Baidu AI child protection products for children use Baidu brain technology to connect with preschool education scenes, and then feed back to the school management system such elements as children and students' listening time, behavior, objects, expressions, and micro actions identified in the classroom. It has achieved a comprehensive assessment of the concentration of children and student groups in the classroom, providing accurate basis for schools and institutions to assess teaching quality and adopt targeted improvement methods. At the same time, the identity of outsiders in the kindergarten, children's dangerous behaviors, negative emotions, and dangerous language semantics were monitored in real time, effectively improving the safety management of the kindergarten.
The flow chart of the above "classroom concentration" solution:
The specific implementation method is as follows:
Step 1: Install multiple cameras (live photos) in front of/at the side of the classroom
The second step: through the comprehensive use of face detection/recognition, human body key point and action recognition and object recognition and other technologies, the captured video stream is sampled and analyzed, and the indicators such as individual classroom concentration, class cumulative concentration and class instantaneous concentration are comprehensively displayed;
Step 3: Install camera/microphone arrays in classrooms and parks to cover and solve the identity detection problem of personnel on campus (live photos).