test.py

# this script should be executed from its parent folder e.g. python ./test.py

import cv2
from keras.models import load_model
import numpy as np
from statistics import mode
from utils import preprocess_input
from utils import get_labels
import sys

# parameters
image_path = './images/test_image.jpg'

detection_model_path = './trained_models/haarcascade_frontalface_default.xml'
emotion_model_path = './trained_models/simple_CNN.530-0.65.hdf5'
gender_model_path = './trained_models/simple_CNN.81-0.96.hdf5'

emotion_labels = get_labels('fer2013')
gender_labels = get_labels('imdb')
font = cv2.FONT_HERSHEY_SIMPLEX

x_offset_emotion = 20
y_offset_emotion = 40
x_offset = 30
y_offset = 60

# loading models
face_detection = cv2.CascadeClassifier(detection_model_path)
emotion_classifier = load_model(emotion_model_path)
gender_classifier = load_model(gender_model_path)

frame = cv2.imread(image_path)
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

faces = face_detection.detectMultiScale(gray, 1.3, 5)
exportCount = 0
for (x,y,w,h) in faces:
    face = frame[(y - y_offset):(y + h + y_offset),
                (x - x_offset):(x + w + x_offset)]

    gray_face = gray[(y - y_offset_emotion):(y + h + y_offset_emotion),
                    (x - x_offset_emotion):(x + w + x_offset_emotion)]
    try:
        face = cv2.resize(face, (48, 48))
        gray_face = cv2.resize(gray_face, (48, 48))
    except:
        continue

    cv2.imwrite('./images/cutouts/' + str(exportCount) + '.png', gray_face)
    exportCount = exportCount + 1

    face = np.expand_dims(face, 0)
    face = preprocess_input(face)

    gender_label_arg = np.argmax(gender_classifier.predict(face))
    gender = gender_labels[gender_label_arg]

    gray_face = preprocess_input(gray_face)
    gray_face = np.expand_dims(gray_face, 0)
    gray_face = np.expand_dims(gray_face, -1)

    #print(len(gray_face))
    #print(len(gray_face[0]))
    #print(len(gray_face[0][0]))
    #print(len(gray_face[0][0][0]))

    emotion_label_arg = np.argmax(emotion_classifier.predict(gray_face))
    emotion = emotion_labels[emotion_label_arg]

    if gender == gender_labels[0]:
        gender_color = (0, 0, 255)
    else:
        gender_color = (255, 0, 0)

    cv2.rectangle(frame, (x, y), (x + w, y + h), gender_color, 2)
    cv2.putText(frame, emotion, (x, y - 90), font,
                    2, gender_color, 2, cv2.LINE_AA)
    cv2.putText(frame, gender, (x , y - 90 + 70), font,
                    2, gender_color, 2, cv2.LINE_AA)

frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
cv2.imwrite('./images/predicted_test_image.png', frame)