Skip to content

This repository provides various Python methods for performing facial prediction.

Notifications You must be signed in to change notification settings

johnbumgarner/facial_prediction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Overview Facial Prediction

The repository Facial Features Detection discussed various Python methods used to identify human faces and multiple facial characteristics.

Primary objective of this repository

This repository is going to examine the OpenCV function face_FaceRecognizer.predict(), which is used to predict the probability that a target photo matches another image within a dataset. This target image is of the well-known female actress Natalie Portman and the dataset contains 73 images of other actresses.

Another objective of this repository is to determine the capabilities and limitations of the Python libraries used to perform the prediction test.

Facial Detection and Facial Prediction

Open Computer Vision Library (OpenCV):

This experiment used the CV2 modules OpenCV-Python and OpenCV-Contrib-Python.. These modules provide functions designed for real-time computer vision, image processing and machine learning.

OpenCV is being used today for a wide range of applications which include:

  • Automated inspection and surveillance
  • Video/image search and retrieval
  • Medical image analysis
  • Criminal investigation
  • Vehicle tag recognition
  • Street view image stitching
  • Robot and driver-less car navigation and control
  • Signature pattern detection on documents

Haar Cascade Classifier - Facial Detection

One of the most basic Haar Cascade classifiers is haarcascade_frontalface_default.xml, whch is used to detect the facial area of a human face looking directly at the camera. This base-level algorithm comes pretrained, so it is able to identify images that have human face characteristics and their associated parameters and ones that have no human face characteristics, such as an image of a cat.

It's worth noting that Python occasionally has issues locating the Haar Cascade classifiers on your system. To solve this you can use the Python module os to find the absolute paths for the classifiers installed.

cv2_base_dir = os.path.dirname(os.path.abspath(cv2.__file__))
haar_frontal_face_model = os.path.join(cv2_base_dir, 'data/haarcascade_frontalface_default.xml')

The Haar Cascade classifier haarcascade_frontalface_default.xml is used in the following matter to located a human face within an image.

# This code was extracted from the script cv2_facial_prediction.py

image_name = 'natalie_portman.jpeg'
photograph = cv2.imread(image_name)
grayscale_image = cv2.cvtColor(photograph, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(grayscale_image, scaleFactor=1.3, minNeighbors=5)
for (x_coordinate, y_coordinate, width, height) in faces:
    cv2.rectangle(photograph, (x_coordinate, y_coordinate),
                  (x_coordinate + width, y_coordinate + height), (255, 0, 255), 2)

The image of Natalie Portman below has a bounding box drawn around the entire facial area identified by the Haar Cascade classifier haarcascade_frontalface_default.xml.

OpenCV Data Gathering

In this phase all the images within the dataset will have their facial parameters calculated using the frontalface Haar Cascade classifier. These data points along with their corresponding identification numbers will be pass to the OpenCV Recognizer for addtional processing.

OpenCV Recognizer Training

In the previous phase facial data was extracted from a dataset of images using the frontal face Haar Cascade classifier. In this phase the data will be run through the OpenCV Recognizer.. The LBPH face recognition algorithm was used in this training phase. The Local Binary Pattern (LBP) is a simple and efficient texture operator which labels the pixels of an image by thresholding the neighborhood of each pixel and considers the result as a binary number.

Parameters: the LBPH uses 4 parameters:

  1. Radius: the radius used for building the Circular Local Binary Pattern. The greater the radius, the smoother the image but more spatial information you can get.

  2. Neighbors: the number of sample points to build a Circular Local Binary Pattern from. An appropriate value is to use 8 sample points. Keep in mind: the more sample points you include, the higher the computational cost. Max value is 8.

  3. Grid X: the number of cells in the horizontal direction, 8 is a common value used in publications. The more cells, the finer the grid, the higher the dimensionality of the resulting feature vector. Max value is 8

  4. Grid Y: The number of cells in the vertical direction, 8 is a common value used in publications. The more cells, the finer the grid, the higher the dimensionality of the resulting feature vector. Max value is 8.

The initial dataset created in the Data Gathering phase contained facial boundary box coordinates, labels and identification numbers. A pickle file named face_labels.pickle will be used to contain the associations between the dataset images and their labels. Pickling is a process where a Python object hierarchy is converted into a byte stream and dumps it into a file by using dump function. This character stream contains all the information necessary to reconstruct the object in another python script.

The facial boundary box coordinates for each image will be processed using the OpenCV LBPHFaceRecognizer. The output will be written to a YML file. This file is primarily associated with Javascript by YAML. YAML stand for "YAML Ain't Markup Language." YAML uses a text file and organizes it into a format which is Human-readable.

# This code was extracted from the script cv2_facial_prediction.py

recognizer = cv2.face.LBPHFaceRecognizer_create(radius=1, neighbors=4, grid_x=4, grid_y=4)

with open('face_labels.pickle', 'wb') as pickle_file:
    pickle.dump(image_names, pickle_file)
 
recognizer.train(x_train, np.array(y_labels))
recognizer.write('face_train_data.yml')

OpenCV Recognizer Prediction

In this phase the data elements within the pickle and YAML files created in the training phase will be used in conjunction with the Haar Cascade classifier haarcascade_frontalface_default.xml. The classifier is used to obtain the boundary box coordinates from the photograph of the unknown person that we are looking for within the image dataset. The OpenCV function face_FaceRecognizer.predict() is used to compute a confidence score, which indicates the match potential between the target image and one within the dataset. A perfect match will have a confidence score of zero. Confidence scores can assigned to various thresholds levels, which will allow for the possibility of close matches and no related matches within the dataset.

# This code was extracted from the script cv2_facial_prediction.py

recognizer.read(training_data)

with open(image_labels, 'rb') as pickle_file:
   pickle_labels = pickle.load(pickle_file)
   labels = {value: key for key, value in pickle_labels.items()}
   
image = cv2.imread(image_name, cv2.IMREAD_UNCHANGED)
image = cv2.resize(image, (image_height, image_width), interpolation=cv2.INTER_AREA)
grayscale_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(grayscale_image, scaleFactor=1.3, minNeighbors=8, flags=cv2.CASCADE_SCALE_IMAGE)

for (x_coordinate, y_coordinate, width, height) in faces:
   cv2.rectangle(image, (x_coordinate, y_coordinate), (x_coordinate + width, y_coordinate + height), (255, 0, 255), 2)
   roi_gray = grayscale_image[y_coordinate:y_coordinate + height, x_coordinate:x_coordinate + width]
   identified_person, confidence_score = recognizer.predict(roi_gray)
   if confidence_score == 0:
      <DO SOMETHING>
   elif 0 < confidence_score <= 10:
      <DO SOMETHING>
   else:
      <DO SOMETHING>

face_FaceRecognizer.predict() was able to accurately predicated that the target image of Natalie Portman matched a face within the dataset that contained 73 photos of well-known female actresses.

The FaceRecognizer algorithm was also able to predicated that the mirrior image of Natalie Portman matched a face within the dataset. The confidence score for this match was 3 precent, which was well within the threshold level for probable matches.

OpenCV Recognizer Prediction Issues:

The dataset used in the prediction test only contained 73 images of well-known actresses. Using this dataset caused some issues when a different photo of Natalie Portman was introduced as the target image. This new target image did not find a probable match within the limited dataset. To compensate for this issue, addtional photos of Natalie Portman were added to a modified dataset to provide more facial possibilities to query.

About

This repository provides various Python methods for performing facial prediction.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages