RetinaFace Python Library

The easiest way to install the library and its prerequisites is via PyPI using pip.

$ pip install retina-face

If you need an end-to-end face recognition pipeline (detect, align, normalize, represent, and verify), it is recommended to use the deepface library. You can specify retinaface as the detector_backend when using models like ArcFace.

#!pip install deepface
from deepface import DeepFace

obj = DeepFace.verify("img1.jpg", "img2.jpg", model_name = 'ArcFace', detector_backend = 'retinaface')
print(obj["verified"])

To perform face alignment (which can increase face recognition accuracy), use RetinaFace.extract_faces(). By setting align = True, the function uses detected facial landmarks to align the faces during extraction.

This returns a list of extracted face images.

import matplotlib.pyplot as plt
from retinaface import RetinaFace

faces = RetinaFace.extract_faces(img_path = "img.jpg", align = True)
for face in faces:
  plt.imshow(face)
  plt.show()

Use RetinaFace.detect_faces() to perform face detection. It requires the exact path to an image as input.

The method returns a dictionary containing facial area coordinates, landmarks (eyes, nose, and mouth), and a confidence score for each detected face.

from retinaface import RetinaFace

resp = RetinaFace.detect_faces("img1.jpg")

The output of detect_faces() is a dictionary where each key is a face identifier (e.g., face_1). Each face object contains:

• score: Confidence score of the detection.
• facial_area: A list containing the coordinates [x1, y1, x2, y2].
• landmarks: A dictionary containing coordinates for right_eye, left_eye, nose, mouth_right, and mouth_left.

{
    "face_1": {
        "score": 0.9993440508842468,
        "facial_area": [155, 81, 434, 443],
        "landmarks": {
          "right_eye": [257.82974, 209.64787],
          "left_eye": [374.93427, 251.78687],
          "nose": [303.4773, 299.91144],
          "mouth_right": [228.37329, 338.73193],
          "mouth_left": [320.21982, 374.58798]
        }
  }
}