Skip to content

Instantly share code, notes, and snippets.

View jonathanoheix's full-sized avatar

jonathanoheix

32 followers · 12 following
  • Macif-Mutualité
  • France
View GitHub Profile
@jonathanoheix
jonathanoheix / main.py
Created January 4, 2019 09:12
from flask import Flask, render_template, Response
from camera import VideoCamera
app = Flask(__name__)
@app.route('/')
def index():
return render_template('index.html')
def gen(camera):
@jonathanoheix
jonathanoheix / model.py
Created January 4, 2019 09:12
from keras.models import model_from_json
import numpy as np
class FacialExpressionModel(object):
EMOTIONS_LIST = ["Angry", "Disgust",
"Fear", "Happy",
"Neutral", "Sad",
"Surprise"]
@jonathanoheix
jonathanoheix / camera.py
Created January 4, 2019 09:11
import cv2
from model import FacialExpressionModel
import numpy as np
facec = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
model = FacialExpressionModel("model.json", "model_weights.h5")
font = cv2.FONT_HERSHEY_SIMPLEX
class VideoCamera(object):
def __init__(self):
@jonathanoheix
jonathanoheix / face8.py
Created January 4, 2019 08:28
# show the confusion matrix of our predictions
# compute predictions
predictions = model.predict_generator(generator=validation_generator)
y_pred = [np.argmax(probas) for probas in predictions]
y_test = validation_generator.classes
class_names = validation_generator.class_indices.keys()
from sklearn.metrics import confusion_matrix
import itertools
@jonathanoheix
jonathanoheix / face7.py
Created January 4, 2019 08:28
# plot the evolution of Loss and Acuracy on the train and validation sets
import matplotlib.pyplot as plt
plt.figure(figsize=(20,10))
plt.subplot(1, 2, 1)
plt.suptitle('Optimizer : Adam', fontsize=10)
plt.ylabel('Loss', fontsize=16)
plt.plot(history.history['loss'], label='Training Loss')
plt.plot(history.history['val_loss'], label='Validation Loss')
@jonathanoheix
jonathanoheix / face6.py
Created January 4, 2019 08:28
# serialize model structure to JSON
model_json = model.to_json()
with open("model.json", "w") as json_file:
json_file.write(model_json)
@jonathanoheix
jonathanoheix / face5.py
Created January 4, 2019 08:28
# number of epochs to train the NN
epochs = 50
from keras.callbacks import ModelCheckpoint
checkpoint = ModelCheckpoint("model_weights.h5", monitor='val_acc', verbose=1, save_best_only=True, mode='max')
callbacks_list = [checkpoint]
history = model.fit_generator(generator=train_generator,
steps_per_epoch=train_generator.n//train_generator.batch_size,
@jonathanoheix
jonathanoheix / face4.py
Created January 4, 2019 08:27
from keras.layers import Dense, Input, Dropout, GlobalAveragePooling2D, Flatten, Conv2D, BatchNormalization, Activation, MaxPooling2D
from keras.models import Model, Sequential
from keras.optimizers import Adam
# number of possible label values
nb_classes = 7
# Initialising the CNN
model = Sequential()
@jonathanoheix
jonathanoheix / face3.py
Created January 4, 2019 08:27
from keras.preprocessing.image import ImageDataGenerator
# number of images to feed into the NN for every batch
batch_size = 128
datagen_train = ImageDataGenerator()
datagen_validation = ImageDataGenerator()
train_generator = datagen_train.flow_from_directory(base_path + "train",
target_size=(pic_size,pic_size),
@jonathanoheix
jonathanoheix / face2.py
Created January 4, 2019 08:27
# count number of train images for each expression
for expression in os.listdir(base_path + "train"):
print(str(len(os.listdir(base_path + "train/" + expression))) + " " + expression + " images")