Attendance system using facial recognition

Project Title

Attendance system using facial recognition

Project Area of Specialization Artificial IntelligenceProject Summary

Our goal is to implement the whole facial recognition system in our university by collecting data from the students and implementing it in classes. It can also work as a security system in which we can deploy a facial recognition system to our gates and allow only those who are registered to come into the university. Our model encodes a face at the backend and extracts 120 features of that face and compares it with the given face at run time. The 120 features are encoded using HOG’s transformation by which comparing similarities is a lot easier and faster as 120 features detect the exact face and the model is trained with time so the speed becomes faster. 

Project Objectives

I and my team am making an attendance system based on facial recognition in which our model will detect the face of the person and mark it by using HOG’s transformation to extract all the features of the face and compare them with the given picture at that time and store the data if matched in a CSV file. 

Project Implementation Method

In the first phase after the software is made the pictures of students in a specific section will be taken. Cameras will be installed in the classroom and the program will be set up. After the success of phase one, we move towards phase two of setting up the camera in every class and collecting data. After this, in the final phase, there will be added an additional feature that will detect if the class is empty or not but the decision of the final phase is still underway.

Benefits of the Project

There is a tremendous scope of facial recognition systems all over the world. The system can be efficiently used in ATMs, driving license verification, identifying duplicate voters, passports via verification, and much more. The benefits of using our system will provide hassle-free attendance without human error.

Technical Details of Final Deliverable

The method we used in this system is HOG’s Transformation. HOG decomposes an image into small squared cells, computes a histogram of oriented gradients in each cell, normalizes the result using a block-wise pattern, and returns a descriptor for each cell. The histogram of oriented gradients (HOG) is a feature descriptor used in computer vision and image processing for the purpose of object detection. We use this feature extraction for our images to compare them with images shown at runtime. It is combined with a linear SVM machine learning algorithm to perform face detection. The SVM machine learning algorithm also known as the SUPPORT VECTOR MACHINE algorithm is used for both classification and regression challenges. An SVM algorithm generates a decision surface separating the two classes. For face recognition, we re-interpret the decision surface to produce a similarity metric between two facial images.

Final Deliverable of the Project Hardware SystemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Quality Education, Affordable and Clean EnergyRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	56000
LAMBDA LABS API	Equipment	1	34000	34000
Camera	Equipment	8	1500	12000
INFERDO API	Equipment	1	10000	10000