Skip to content

Latest commit

 

History

History
53 lines (39 loc) · 2.2 KB

README.md

File metadata and controls

53 lines (39 loc) · 2.2 KB
Raw

Face Detection Using Facenet

Project for IITISoC

This project performs Multi-person face detection using MTCNN, while the main model is trained using Siamese Network(Facenet).
The front-end is developed from Flask WebApp.

Due to the version of dependencies required to build this project, the version of Python should mandatorily be 3.6.8.

Steps to run:

  1. Install dependencies from requirements_cpu.txt or requirements_gpu.txt using the following code:

    pip install -r requirements_cpu.txt

    or

    pip install -r requirements_gpu.txt

    Note: If you have a NVIDIA GPU, you must first install the CUDA libraries(Link-http://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/index.html).
    Note: It is advised to create this project on a python virtual environment(Link-https://docs.python.org/3/tutorial/venv.html).

  2. Install the pre-trained model from the link-https://drive.google.com/file/d/0B5MzpY9kBtDVZ2RpVDYwWmxoSUk/edit.
    Place this model under the model folder such that the path of the model corresponds as:
    ../model/20170512-110547/20170512-110547.pb

  3. Run the flask server with the following code:

    python server.py

    The server is set up on default URL:localhost:5000

  4. Upload your images using the app interface(.png,.jpeg,.jpg only allowed). Make sure to name the image file as the name of the person inside the image. Note:When the image file is uploaded successfully, the cropped face images will appear in the uploads/ folder, and the corresponding embedding files will appear in the embeddings/ folder.

  5. Click on Click here for live facial recognition with Web Camera! button to start the webcam feed. Press q from keyboard to take the image and close the webcam feed.

  6. Voilà, you'll get the all the present individuals displayed right on the image taken from Webcam.

References: Facenet and MTCNN model source:"https://github.com/davidsandberg/facenet"

Team:
Aryan Rastogi
Ashish Gautam
Aditya Bharadwaj
Bharat Gupta
Vardhan Paliwal
Purnadip Chakrabarti

This project was created for IITISoC 2020.