def main():
root_dir_train = "./Task_1/"
root_dir_test = "./Task_1/"
train_path = os.path.join(root_dir_train, "development/") # train images
test_path = os.path.join(root_dir_test, "evaluation/") # train images
create_csv(root_dir_train, root_dir_test, train_path, test_path)
preprocess = FaceRecog(margin=7)
batch_size = 55
# model = models.resnet18(pretrained=False)
model = models.resnet34(pretrained=False)
clssf = BinaryClassifier(model, freeze=False)
train_data = Task1_loader("./Task_1/train.csv",
phase='train',
preprocess=preprocess)
test_data = Task1_loader("./Task_1/test.csv",
phase='test',
preprocess=preprocess)
train_loader = DataLoader(train_data,
batch_size=batch_size,
shuffle=True,
num_workers=8)
valid_loader = DataLoader(test_data,
batch_size=batch_size,
shuffle=False,
num_workers=8)
criterion = nn.BCELoss()
# optimizer = optim.SGD(model.parameters(), lr=0.0018, momentum=0.27)
optimizer = optim.Adam(model.parameters(), lr=0.0018, weight_decay=0.0015)
train(clssf,
train_loader,
valid_loader,
criterion,
optimizer,
10,
device='cpu')
def main():
root_dir_train = "./Task_1/"
root_dir_test = "./Task_2_3/"
train_path = os.path.join(root_dir_train, "development/") # train images
test_path = os.path.join(root_dir_test, "evaluation/") # train images
create_csv(root_dir_train, root_dir_test, train_path, test_path)
preprocess = FaceRecog(margin=7)
batch_size = 35
model = models.resnet34(pretrained=False)
clssf = BinaryClassifier(model, freeze=False)
train_data = Task1_loader(root_dir_train + "train.csv",
phase='train',
preprocess=preprocess)
test_data = Task1_loader(root_dir_test + "test.csv",
phase='test',
preprocess=preprocess)
train_loader = DataLoader(train_data,
batch_size=batch_size,
shuffle=True,
num_workers=8)
valid_loader = DataLoader(test_data,
batch_size=batch_size,
shuffle=False,
num_workers=8)
criterion = nn.BCELoss()
optimizer = optim.Adam(model.parameters(), lr=0.0018, weight_decay=0.0015)
load_checkpoint(model, 'checkpoints/task1_84_7.pkl', optimizer)
valid_loss, accuracy = validate(model, valid_loader, criterion, 'cpu')
print(f'val loss: {valid_loss:04f} ' f'val acc: {valid_acc*100:.4f}%')
def main():
root_dir_train = "./Task_1/"
root_dir_test = "./Task_2_3/"
train_path = os.path.join(root_dir_train, "development/") # train images
test_path = os.path.join(root_dir_test, "evaluation/") # train images
create_csv(root_dir_train, root_dir_test, train_path, test_path)
# preprocess()
batch_size = 35
model = models.resnet50(pretrained=True)
clssf = BinaryClassifier(model)
train_data = Task1_loader(root_dir_train + "train.csv", phase='train')
test_data = Task1_loader(root_dir_test + "test.csv", phase='test')
train_loader = DataLoader(train_data,
batch_size=batch_size,
shuffle=True,
num_workers=8)
valid_loader = DataLoader(test_data,
batch_size=batch_size,
shuffle=False,
num_workers=8)
#criterion = nn.CrossEntropyLoss()
criterion = nn.BCELoss()
optimizer = optim.SGD(model.parameters(), lr=0.001, momentum=0.2)
# optimizer = optim.SGD(model.parameters(), lr=0.0015, momentum=0.26)
train(clssf,
train_loader,
valid_loader,
criterion,
optimizer,
10,
device='cpu')
'Missing arguments, you should set:\
- The model unique identifier\
- The directory to store all generated outputs\
- The training file\
Received: ' + ' '.join(sys.argv)
]
# Add the provided unique id.
if len(sys.argv) > 1:
result['modelid'] = sys.argv[1]
print(json.dumps(result))
sys.exit(result['status'])
modelid = sys.argv[1]
directory = sys.argv[2]
# Sklearn binary classifier - logistic regression.
binary_classifier = BinaryClassifier(modelid, directory)
# TensorFlow binary classifier - NN.
#binary_classifier = BinaryClassifierTensorFlow(modelid, directory)
# TensorFlow binary classifier - logistic regression.
#binary_classifier = BinaryClassifierSkflow(modelid, directory)
# TensorFlow binary classifier - deep neural network.
#binary_classifier = BinaryClassifierDNN(modelid, directory)
result = binary_classifier.train_dataset(sys.argv[3])
print(json.dumps(result))
sys.exit(result['status'])
- The minimum deviation to accept the model as valid (defaults to 0.02)\
- The number of times the evaluation will run (defaults to 100)\
Received: ' + ' '.join(sys.argv)
]
# Add the provided unique id.
if len(sys.argv) > 1:
result['modelid'] = sys.argv[1]
print(json.dumps(result))
sys.exit(result['status'])
modelid = sys.argv[1]
directory = sys.argv[2]
# Sklearn binary classifier - logistic regression.
binary_classifier = BinaryClassifier(modelid, directory)
# TensorFlow binary classifier - NN.
#binary_classifier = BinaryClassifierTensorFlow(modelid, directory)
# TensorFlow binary classifier - logistic regression.
#binary_classifier = BinaryClassifierSkflow(modelid, directory)
# TensorFlow binary classifier - deep neural network.
#binary_classifier = BinaryClassifierDNN(modelid, directory)
result = binary_classifier.evaluate_dataset(sys.argv[3], float(sys.argv[4]),
float(sys.argv[5]),
int(sys.argv[6]))
print(json.dumps(result))
sys.exit(result['status'])
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from BinaryClassifier import BinaryClassifier
train = pd.read_csv("./Dataset/train.csv").drop(["PassengerId", "Name", "Ticket", "Embarked", "Cabin"], axis=1)
test = pd.read_csv("./Dataset/test.csv").drop(["PassengerId", "Name", "Ticket", "Embarked", "Cabin"], axis=1)
train_labels = train["Survived"]
def gender_to_int(sample): return int(sample == "male")
train_gender = [[gender_to_int(sample)] for sample in train["Sex"].values]
test_gender = [[gender_to_int(sample)] for sample in test["Sex"].values]
train_features = np.concatenate([train.drop(["Sex"], axis=1).values, train_gender], axis=1)
test_features = np.concatenate([test.drop(["Sex"], axis=1).values, test_gender], axis=1)
train_ds = tf.data.Dataset.from_tensor_slices((train_features, train_labels)).shuffle(100)
val_ds = train_ds.take(200).batch(200)
train_ds = train_ds.skip(200).batch(64)
test_ds = tf.data.Dataset.from_tensor_slices((test_features)).batch(64)
model = BinaryClassifier([8, 4, 1])
model.compile(tf.keras.optimizers.Adadelta(0.0001), tf.keras.losses.MeanSquaredError(), [tf.keras.metrics.Accuracy()])
model.fit(train_ds.take(1), epochs=20, verbose=2, validation_data=val_ds)
#for sample in train_ds.take(1):
# print(sample)
TEST_IMAGES = ['./images/5.jpg', './images/6.jpg']
TEST_TRUTHS = ['./images/5_mask.jpg', './images/6_mask.jpg']
# Change this to run a different pipeline
RUN_PIPELINE = 2
if RUN_PIPELINE == 1:
# Pipeline 1
## This trains on the train images with 3 gaussians, a threshold of 0.5 and then
## tests the model on the test images.
print("Running pipeline 1")
APPLE_CLASSIFIER = BinaryClassifier(gabor=False, log=True)
APPLE_CLASSIFIER.load_train_images(TRAIN_IMAGES,
TRAIN_TRUTHS,
display=False)
APPLE_CLASSIFIER.train(k=3)
APPLE_CLASSIFIER.set_threshold(0.5)
APPLE_CLASSIFIER.test(TEST_IMAGES, truths=TEST_TRUTHS, display=True)
elif RUN_PIPELINE == 2:
# Pipeline 2
## This tunes the model over a variable number of gaussians (max 5) and threshold
## (between 0.01 and 0.99 in steps of 0.01), finds the optimum parameters and
## trains the model on these. Finally, the model is tested against the test images.
print("Running pipeline 2")