Coverage for verification/face_verify.py : 82%

import dlib 

from scipy.spatial import distance 

from hutts_utils.hutts_logger import logger 

class FaceVerify: 

""" 

The FaceVerify class is responsible for 

1. Detecting the face. 

2. Generating a threshold value that determines the likeness 

of two individuals in an image 

""" 

def __init__(self, shape_predictor_path, face_recognition_path): 

""" 

Initialise face verify manager 

Args: 

shape_predictor_path (str): Describes the path the Shape Predictor 

trained data. 

face_recognition_path (str): Describes the path the face recognition 

trained data. 

Raises: 

TypeError: If a none string value is passed for shape_predictor_path 

TypeError: If a none string value is passed for face_recognition_path 

""" 

logger.info("Initialise FaceVerify") 

if not isinstance(shape_predictor_path, str): 

raise TypeError( 

'Bad type for arg shape_predictor_path - expected string. Received type "%s".' % 

type(shape_predictor_path).__name__ 

) 

if not isinstance(face_recognition_path, str): 

raise TypeError( 

'Bad type for arg face_recognition_path - expected string. Received type "%s".' % 

type(face_recognition_path).__name__ 

) 

self.shape_predictor_path = shape_predictor_path 

self.face_recognition_path = face_recognition_path 

def verify(self, face1, face2, threshold=0.55): 

""" 

This function determines a percentage value of how close the faces 

in the images passed are to each other if the determined value if below 

the threshold value passed by the user a boolean value of True is returned 

indicating that the faces in the images passed indeed match. 

The Verify function makes use of the dlib library which guarantees 99.38% 

accuracy on the standard Labeled Faces in the Wild benchmark. 

Author(s): 

Stephan Nell 

Args: 

face1 (:obj:'OpenCV image'): The first image containing the face that should be compared. 

face2 (:obj:'OpenCV image'): The second image containing the face that should be compared 

threshold (float): The threshold value determines at what distance the two images are 

considered the same person. If a verify score is below the threshold value the faces are 

considered a match. The Labled Faces in the Wild benchmark recommend a default threshold 

of 0.6 but a threshold of 0.55 was decided on since a threshold of 0.55 represents 

the problem better. 

Returns: 

bool: Represent if two face indeed match True if distance calculated is 

below threshold value. False if the distance calculated is 

above threshold value. 

float: Return Euclidean distance between the vector representation 

of the two faces 

Raises: 

ValueError: If no face can be detected no faces can be matched and 

operation should be aborted. 

""" 

logger.debug('Getting frontal face detector') 

detector = dlib.get_frontal_face_detector() 

logger.debug('Getting shape predictor') 

shape_predictor = dlib.shape_predictor(self.shape_predictor_path) 

logger.debug('Getting facial recogniser') 

facial_recogniser = dlib.face_recognition_model_v1(self.face_recognition_path) 

logger.info('Getting face in first image') 

face_detections = detector(face1, 1) 

if len(face_detections) == 0: 

logger.error('Could not find a face in the first image') 

raise ValueError('Face could not be detected') 

logger.debug('Getting the shape') 

shape = shape_predictor(face1, face_detections[0]) 

logger.debug('Getting the first face descriptor') 

face_descriptor1 = facial_recogniser.compute_face_descriptor(face1, shape) 

logger.info('Getting face in second image') 

face_detections = detector(face2, 1) 

if len(face_detections) == 0: 

logger.error('Could not find a face in the first image') 

raise ValueError('Face could not be detected') 

logger.debug('Getting the shape') 

shape = shape_predictor(face2, face_detections[0]) 

logger.debug('Getting the second face descriptor') 

face_descriptor2 = facial_recogniser.compute_face_descriptor(face2, shape) 

logger.info('Calculating the euclidean distance between the two faces') 

match_distance = distance.euclidean(face_descriptor1, face_descriptor2) 

logger.info('Matching distance: ' + str(match_distance)) 

# Any distance below our threshold of 0.55 is a very good match. 

# We map 0.55 to 85% and 0 to 100%. 

if match_distance < threshold: 

match_distance = 1 - match_distance 

threshold = 1 - threshold + 0.05 

percentage_match = ((match_distance-threshold)*15/50)*100 + 85 

logger.info('Matching percentage: ' + str(percentage_match) + "%") 

return True, percentage_match 

113 ↛ 115line 113 didn't jump to line 115, because the condition on line 113 was never true elif match_distance < threshold + 0.05: 

# In this if we map (0.55-0.60] we map 0.549 to 70% match 

match_distance = 1 - match_distance 

threshold = 1 - threshold + 0.05 

percentage_match = ((match_distance-threshold)*30/55)*100 + 70 

logger.info('Matching percentage: ' + str(percentage_match) + "%") 

return True, percentage_match 

else: 

# If the distance is higher than 0.65 we map it to 60% and below 

percentage_match = 60 - ((match_distance-threshold)*60/40)*100 

logger.info('Matching percentage: ' + str(percentage_match) + "%") 

return False, percentage_match