Facial-Recognition-Project
-
Credits: Adam Geitgey
-
The main directory is face-identification-project but it is preferable to understand facial landmark detection first in order to execute the recognition part.
This project is completed using face_recognition package. This package is compatible with DLIB Library.
- There are two jupyter notebooks in the face-identification directory:
face-identification [Built on Google Colab]
- Advantages on Colab:
-
GPU: Greater computing power, therefore better experience and faster results.
-
No need for Dependencies Installment: In colab there is no need of installing dependencies.
- Disadvantages:
-
Quite rigid for OpenCV: There are certain restrictions in Colab while using OpenCV.
-
No webcam Control: No alternate of controlling webcam using OpenCV in google colab, although you can just capture an image.
Face-Identification [Built on Local Machine]
- Disadvantages on Local Machine:
-
CPU: Less computing power therefore, “cnn” model is to be replaced with “hog”
-
Dependencies Installment: Packages need to install: OpenCV, face_recognition and also face_recognition is not officialy compatible with Windows but there ways to install it.
-
Advantages:
-
Once everything is installed the process is smooth.
SAMPLE OUTPUTS:
Installing face_recognition package on Windows machine:
For working on Local Machine
Step1: Download and install python 3.8 or higher (64 bit)
-
visit Python Downloads and
Download Windows x86-64 executable installer. -
Note:
1. 32 bit for this project is useless.
2. While installing python tick ADD to PATH checkbox
Step2: Installing Microsoft Build Tools
Step 3: pip install cmake.
Step3: Installing face_recognition Package
There is no need to install Dlib Explicitly as face_recognition package comes along with this.
This is the best Video Tutorial I have seen.
Facial Attendance System
- Logic: Facial Recognition Attendance System: Recognizing the face and checking whether the name is present in the CSV or Not. If yes then it does not overwrite and if no it will write the name with date.
File Structure:
- Facial-Attendance-System:
- detect.py: Main script
- Attendance.csv : Attendance wil be marked here.
- Known_faces : Consists the faces of people that needs to be recognised.