A Chinese School is Using Facial Recognition to Analyze Students’ Behavior

Pranav Dar 10 May, 2019 • 3 min read


  • A school in Hangzhou, China, is using facial recognition to monitor the behavior of their students
  • The technology classifies the students based on their range of emotions – from antipathy to happy (and a whole host of others)
  • The system also cross-checks the faces of all students against the school database to mark the attendance and has the ability to predict if a student is feeling sick



The perils of AI are well documented. You must have seen Elon Musk preaching that it has a lot of downsides as well as upsides. But when it comes to using it to monitor the behavior of people, which side of the line do you stand on – do you think this is beneficial or intrusive? China has been emphatic in using AI to transform its way of living and they are now using it in education as well.

Source: AP Photo/Mark Schiefelbein

A school in Hangzhou, China, is using machine learning in the form of facial recognition to monitor and analyze how students are behaving and responding to the teacher. The technology is built into cameras and is called “smart eyes”. It aims to provide real-time data to the teachers (and supervisors) on the emotions their students are displaying in the class. The underlying concept behind this system analyses the facial expression of the students and classifies them into the below categories:

  • Surprised
  • Sad
  • Angry
  • Happy
  • Antipathy
  • Afraid
  • Neutral

The technology has also been programmed to record various actions students perform – like reading, writing, raising their hands, and even if a student falls asleep. This not only helps understand which students are slacking, it is also helping teachers alter and tailor their teaching style to get a more positive response.

Another use case of this system is to monitor the attendance of students by checking students’ faces with the records it has in the school database. The system also manages to detect whether a student is not feeling well and flags it to the teacher.

Unfortunately the results of most actions have not yet been published. How does the school react when a student is flagged as sad or angry? What if a student falls asleep? These are just some of the questions that are yet to be answered. But one thing is for certain, Chine is miles ahead of the rest when it comes to using facial recognition technology to help and monitor it’s citizens.

Watch the video below to see the technology in action:


Our take on this

Using face recognition for scanning behaviour of students in an interesting approach but using the systems in schools has raised obvious privacy concerns. There are upsides to this as well though – it helps the school tailor it’s approach better to keep the students involved.

China has been using facial recognition technology for various other applications as well, like predicting a crime before it happens. Other than China, countries like the US have also used facial recognition technology in schools to protect children. From the perspective of machine learning, this is a step forward towards AI being all pervasive. What is your take on this latest development? Does it excite or scare you? Let us know in the comments below.


Subscribe to AVBytes here to get regular data science, machine learning and AI updates in your inbox!


Pranav Dar 10 May 2019

Senior Editor at Analytics Vidhya.Data visualization practitioner who loves reading and delving deeper into the data science and machine learning arts. Always looking for new ways to improve processes using ML and AI.

Frequently Asked Questions

Lorem ipsum dolor sit amet, consectetur adipiscing elit,

Responses From Readers


Pranov Shobhan Mishra
Pranov Shobhan Mishra 02 Jun, 2018

This is a great and exciting technique and has tremendous potential of automating a lot of non value added activities. In schools , colleges etc, this should be used to get the attendance details instead of teachers wasting time on calling out the names/roll# for attendance. In bigger classes, students do proxy voting for their friends and the purpose of taking attendance is missed. Facial recognition should be installed and manual attendance taken out of equation. Similarly, I had recommended in an office bus transport system to replace the practice of people providing their attendance through an app when they get in, to an automated attendance tracking method through the facial recognition algorithm implementation. This would have helped realtime tracking of how many passengers are in which nodal point and plan accordingly for bus arrangements. I guess, OLA and Uber could do that as well and do away with the need for drivers feed in entry of their customer.The use cases can be immense. I can just go on and on.

Yatin Jain
Yatin Jain 08 Oct, 2018

Hi Pranav,This technology seems very interesting to me.Can you guide if I want to bring this technology in India from where I can buy and understand the usability of this. Also if you can guide me the approx cost of installing this or any other details.

Bruce 06 Nov, 2018

Facial recognition systems do not keep the face images - they reduce the face to a set of numbers (a vector), which is then stored. There is no way to recreate the facial image from those vectors. At least with the system I work on.And those vectors only have meaning to that application.Thus, there is limited value in a database of vectors - which should be encrypted anyway, like any database.

s 01 Jun, 2023

It's a good idea and they will find a way to implement it properly.

  • [tta_listen_btn class="listen"]