facial expression and Recognition using deep Learning

Project Title

facial expression and Recognition using deep Learning

Project Area of Specialization Artificial IntelligenceProject Summary

In classical teaching and classrooms, the quality of the teaching is measured by using surveys collected after the course completion or after the lecture by comparing students’ performance or achievement on the related exams. These approaches can help the presenter or lecturer to make certain changes in lecturing style but these cannot facilitate implementation of improvement during the lecture. Smart classroom system is the type of the system which enables any lecturer to monitor the current level of interest of the present audience or students in the classroom.

The proposed methodology makes the environment of classroom very interactive by integrating various machine learning techniques. The classroom is equipped with two cameras, one facing the students from the front, we name it C1, and one facing the lecturer from the bottom, we name it C2. The C1 will record the student activity and extract the faces from the live stream. The detection of faces is very much matured in literature, however, detecting the faces from the live stream is still very challenging and requires research attention. Once the stable faces are detected, these faces are given to already trained classifiers to identify the mental state of the students by analyzing the facial expressions using computer vision and deep learning approaches.

The C2 will record teacher and lecture slides. Teacher will be holding switch key to control the focus of the C2. It is seen in e-learning videos that the lecture recording is done by some expert which process the video after the lecture and compile it. In our proposed framework, this video will be recorded by our automated program and will be available instantly to audience for e-learning.

Project Objectives

The objective of the projects are as follow:

• Installation of the Cameras in the classroom to make it model classrooms
• Recording of the lectures for facial expression learning
• Training a deep network of facial expressions
• Designing the prototype for lecturer to analyze the class quality by students facial expression 

Project Implementation Method

Project will be designed and implemented using python and will be deployed using dotnet. Following tools will be used for prototyping and classifier

Language:

Python

C#

Libraries:

• Anaconda
• Tensorflow
• Keras
• Java scripts

Hardware required:

• NVR x 1
  • For video recording and monitering
• Camera x 2
  • IP based cameras which will be installed inside the classroom and connected to the server

Benefits of the Project

Through smart class room system teacher can monitor the current level of interest of students.

Teachers can know about their quality of teaching during lecture this will help teachers to improve their teaching method during class lecture

Recorded lecture which will be available for e-learning. which helps students to learn in more efficient way.

Technical Details of Final Deliverable

• Camera C1 facing students from the front. using computer vision and deep learning approaches it identify faces and detect the mental state. through this teacher monitor the current level of interest of students during lecture in class room .it helps to improve quality of method.
• Camera C2  will record the teacher and lecture slides ,the video will be  recorded by our automated program . and these recorded lectures will be available for e learning.

Final Deliverable of the Project HW/SW integrated systemType of Industry Education Technologies Artificial Intelligence(AI)Sustainable Development Goals Quality EducationRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	69000
IP Camera	Equipment	2	18000	36000
NVR	Equipment	1	18000	18000
Switch (POS)	Equipment	1	9000	9000
Camera Stand, installations	Equipment	1	6000	6000