Face Recognition Details 

Detects fraudulent UPI transactions in real time using a machine learning–based classification model. 

Project Description 

This project implements a face recognition system using the K-Nearest Neighbors (KNN) algorithm to identify individuals from images. Facial features are extracted using the face_recognition and dlib libraries, which generate unique face encodings for each person. The system compares these encodings with stored data to recognize faces accurately. A simple interface allows users to upload images and get recognition results, making it suitable for academic and real-world demonstrations. 

Features 

• Face detection and recognition using KNN algorithm

• Facial feature extraction using face_recognition and dlib

• Accurate identification based on facial encodings

• Supports training on multiple known faces

• Image upload–based recognition workflow

• Clean and well-structured implementation

• Suitable for academic, demo, and final-year projects

Tech Stack

• Python

• K-Nearest Neighbors (KNN)

• face_recognition library

• dlib

• Pandas

• Numpy

• OpenCV (for image handling)

• Flask 

• HTML

• JavaScript

System Requirements

• Python 3.10+

• Anaconda

• Libraries: numpy, pandas, scikit-learn, matplotlib

• No GPU required

Deliverables 

• Jupyter Notebook

• Dataset

• Setup over video call

• WhatsApp support