Face-Recognition and Attendance

Learned and implemented the Face detection and recognition using face_recognition api

Attendance

Using face recognition api, self implemented algo in SceneChangeDetect.py and Multithreading pool developed a system in which one can pass argument as past cctv footage and can do facial_recognition(attendance).

Processing Speed:

50sec(changes rapid) -> 10 sec processing and largely depends upon the visual it is getting as it skips most of the part of frame in which no change occured and also no face detected.

Using Attendance system

1. Clone the repository into a python virtual environment with installed packages :

   • face_recognition
   • dlib
   • imutils
   • argparse
   • opencv-python
   • pickle

     clone :

     $ git clone https://github.com/Akash16s/Face-Recognition-Attendance.git

   Structure of files :

    -dataset
         - Name of person or ID
              - Photos
    -input CCTV video

   Now first create encodings.pickle file having encodings of faces

   (venv)$ python encode.py

   Now we can perform the Attendance part

   (vnev)$ python attendance.py -i "Location of input Video" -d "Input the date and time %Y-%m-%d %H:%M:%S"