In this blog, I will tell you how I and my classmate, Het Gohil, made an AI-enabled surveillance system using CNN. For those of you, who are not aware of CNN, let's talk about that first.
Convolutional neural networks (CNNs) are a class of Artificial neural networks(ANNs). They are one of the most powerful and popular tools for machine learning and data science. They can perform tasks such as image recognition, natural language processing, face detection, and more with high accuracy and efficiency. They are inspired by the human visual system and are adept at processing visual information in a hierarchical and systematic manner.
ABOUT OUR PROJECT
The primary goal of our project was to develop a robust and efficient facial recognition system using state-of-the-art CNN techniques. By employing CNN, we aimed to enhance the accuracy, speed, and versatility of facial recognition systems. We incorporated real-time processing and adaptability to continually improve recognition abilities over time to make our software truly efficient. We also prioritized user privacy and data security in our project.
ABOUT THE CODE
This was the first time I was using CNN in Python. We educated ourselves about CNN by reading the blogs available on the Web. Our teacher, Mr. Alpesh Shah, also guided us in implementing what we had read. We learned the basics of CNNs, such as input layers, convolutional layers, activation layers, pooling layers, flattened and fully connected layers, and output layers functions. We coded our first CNN in Python by using the facial recognition library. Face encoding is a way to represent the face using a set of 128 computer-generated measurements. We prepared our data by inputting pictures of known persons. The code underwent a pre-processing stage where the facial images were normalized and resized. We then trained our model by feeding other new faces and letting the model compare them with the known faces. We then evaluated our results. Now, we needed to fine-tune our parameters. We made changes to the resolution and color of the input image to make our model more efficient.
WHAT I GAINED
This project was quite challenging. It took up a lot of our time, which, everyone knows is a limited resource when you are in high school. Still, it was really fun and stimulating to learn a new deep-learning skill. It has boosted my confidence in many ways and opened up a whole world to me.
Reference:
https://www.lystloc.com/blog/7-new-facial-recognition-technology-trends-to-boom-in-the-future/#:~:text=Systems%20for%20access%20control%20are,to%20improve%20security%20and%20privacy
https://thinkingneuron.com/face-recognition-using-deep-learning-cnn-in-python/
https://docs.opencv.org/3.4/da/d60/tutorial_face_main.html
Comments
Post a Comment