-
-
home page
-
login page
-
signup page
-
Teacher's Dashboard
-
update test details
-
Student's dashboard
-
test environment loading
-
successful setup of test environment and detection of student peeping.
-
Real time detection while test
-
auto submission if student tries to exit full screen
Inspiration
I asked my class representative about how exactly exams are going to be conducted and her reply dismayed me, she replied that even professors don't have any clue or any idea how they are going to conduct examination amid COVID-19.I took it as a challenge to build a platform for online examination with an advanced monitoring system.
What it does
It is an online examination platform which ensures the strictness of exam in a virtual environment. It has a multi-tier invigilation system and it involves our smart trained AI which informs the exam-invigilator(human) about the suspected student. That makes the job easy for the conduction of online examination. Below are the details and some touches of our platform:
Teachers and students have a separate portal. In the teacher's section, they can add a test and evaluate the answer sheets. In the student's section, they can take a test just by entering the test-id and test-password given by their teacher As soon as you enter the test we will ask you to give access to your webcam and stabilize yourself in your sitting location and when everything is fine you can enter the test in a click of a button. The test will open in a popup window which should be kept in a full-screen and the test starts on time. As you are in full-screen you can never escape from it else test will be submitted. During the test, we have used state of the art Head-pose detection algorithm along with face detection which makes it impossible for any student to do any malpractices and ensures that there is no other person allowed in the room. Additionally, we have also thought of active-IP detection so, that same account can't be logged in from various devices at the same time. Illegal access to URL manipulation has been taken care of using the intelligent use of session and database updates.
How we built it
We have used various technologies to build this project. A brief description - Front end - HTML5, CSS, BOOTSTRAP Backend - Javascript Server - Flask (python) database - MYSQL Invigilation - We have used state of the art Head-pose detection algorithm along with face detection.
Challenges we ran into
We faced many challenges in this project(As this was our biggest project). The biggest challenge was using OpenCV and face-API to make our thinking come to action, it was really difficult to estimate the landmarks of the face objects and taking the decision on what should be the maximum head deviation and how frequently it should deviate . Another challenge we faced was to ensure that the test taker cannot switch tabs in the browser or exit fullscreen. Giving a pleasant design to the website was also a big task as none of us were having prior experience in web design.
Deploying was also crucial in order to showcase our work to everyone, again we haven't done this before so it was also a challenge. the list goes on but we didn't lose hope any single moment and things got sorted out.
Accomplishments that we're proud of
Learning a lot of techs for first time and implementing them was not an easy task but we have done it. In the end, we have made this full-fledged project which is a working project. We gained some popularity and appreciation.
We are proud of the spirit that at least we have tried to make something that would be useful in these times of crisis. Most important, we are proud of each other who made this possible.
What I learned
We learned a lot while building this platform, we dealt with creation of complex and secure login and signup module, head pose estimation in browser. We dealt with image processing with openCV and the most important part that is deploying our platform that can be used of everyone.
What's next for Anvix - Online examination portal
Things that have to be added:- Invigilator monitoring system which gets the live video of test-taking student and our system lets the invigilator know which student is probably using unfair means that will help them in implementing stricter monitoring which can be accessed directly from the teacher's dashboard by entering invigilator's password that they have assigned at the time of test creation.
And after completion of this, we aim to launch it as a platform for institutes to solve this problem of online examination and proctoring.
Log in or sign up for Devpost to join the conversation.