Face Recognition
Project description
Face Recognition
Simple library to recognize faces from given images
Face Recognition pipeline
Below the pipeline for face recognition:
- Face Detection: the MTCNN algorithm is used to do face detection
- Face Alignement Align face by eyes line
- Face Encoding Extract encoding from face using FaceNet
- Face Classification Classify face via eculidean distrances between face encodings
How to install
pip install git+https://github.com/paoloripamonti/face-recognition
How to train custom model
Initialize model
from face_recognition import FaceRecognition
fr = FaceRecognition()
Train model with pandas DataFrame:
fr = FaceRecognition()
fr.fit_from_dataframe(df)
where 'df' is pandas DataFrame with column person (person name) and column path (image path)
Train model with folder:
fr = FaceRecognition()
fr.fit('/path/root/')
the root folder must have the following structure:
root\
Person_1\
image.jpg
...
image.jpg
Person_2\
image.jpg
...
image.jpg
...
Save and load model
you can save and load model as pickle file.
fr.save('model.pkl')
fr = FaceRecognition()
fr.load('model.pkl')
Predict image
fr.predict('/path/image.jpg')
Recognize faces from given image. The output is a JSON with folling structure:
{
"frame": "image data", # base64 image with bounding boxes
"elapsed_time": time, # elapsed time in seconds
"predictions": [
{
"person": "Person", # person name
"confidence": float, # prediction confidence
"box": (x1, y1, x2, y2) # face bounding box
}
]
}
Example
For more details you can see: https://www.kaggle.com/paoloripamonti/face-recogniton
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
File details
Details for the file facenet-face-recognition-0.1.tar.gz
.
File metadata
- Download URL: facenet-face-recognition-0.1.tar.gz
- Upload date:
- Size: 5.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.36.1 CPython/3.7.0
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
bf741943552be03b87acb8c15cc14d3ade4ca491a93de56981100a98d9e59398
|
|
MD5 |
3c63b36e1d03a6b375afbede7374dcbb
|
|
BLAKE2b-256 |
eb6066bbdea11f97e9d39c2df91d57c6b7ae4f748ccb95206d60a32722f0d550
|