mirror of
https://github.com/serengil/deepface.git
synced 2025-06-07 12:05:22 +00:00
test
This commit is contained in:
Onur Atakan ULUSOY
GitHub
parent
4cc2fd8782
commit
84a1ac64a5
@ -1,353 +1,223 @@
|
|||||||
import warnings
|
import warnings
|
||||||
warnings.filterwarnings("ignore")
|
|
||||||
|
|
||||||
import os
|
import os
|
||||||
#os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
|
import tensorflow as tf
|
||||||
|
import cv2
|
||||||
|
from deepface import DeepFace
|
||||||
|
|
||||||
|
print("-----------------------------------------")
|
||||||
|
|
||||||
|
warnings.filterwarnings("ignore")
|
||||||
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
|
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
|
||||||
|
|
||||||
import json
|
tf_major_version = int(tf.__version__.split(".")[0])
|
||||||
import time
|
|
||||||
import unittest
|
|
||||||
|
|
||||||
#-----------------------------------------
|
if tf_major_version == 2:
|
||||||
|
import logging
|
||||||
|
tf.get_logger().setLevel(logging.ERROR)
|
||||||
|
|
||||||
import tensorflow as tf
|
print("Running unit tests for TF ", tf.__version__)
|
||||||
|
|
||||||
class deepface_unit_tests(unittest.TestCase):
|
print("-----------------------------------------")
|
||||||
|
|
||||||
def test_deepface(self):
|
expected_coverage = 97
|
||||||
tf_version = int(tf.__version__.split(".")[0])
|
num_cases = 0; succeed_cases = 0
|
||||||
|
|
||||||
if tf_version == 2:
|
def evaluate(condition):
|
||||||
import logging
|
|
||||||
tf.get_logger().setLevel(logging.ERROR)
|
|
||||||
|
|
||||||
print("Running unit tests for TF ", tf.__version__)
|
global num_cases, succeed_cases
|
||||||
|
|
||||||
from deepface import DeepFace
|
if condition is True:
|
||||||
from deepface.commons import functions
|
succeed_cases += 1
|
||||||
from deepface.basemodels import VGGFace, OpenFace, Facenet, FbDeepFace
|
|
||||||
from deepface.extendedmodels import Age, Gender, Race, Emotion
|
num_cases += 1
|
||||||
|
|
||||||
print("-----------------------------------------")
|
# ------------------------------------------------
|
||||||
#-----------------------------------------
|
|
||||||
|
|
||||||
print("DeepFace.detectFace test")
|
detectors = ['opencv', 'mtcnn', 'retinaface']
|
||||||
#detectors = ['opencv', 'ssd', 'dlib', 'mtcnn', 'retinaface']
|
models = ['VGG-Face', 'Facenet', 'Facenet512', 'ArcFace', 'SFace']
|
||||||
detectors = ['opencv', 'ssd', 'mtcnn', 'retinaface']
|
metrics = ['cosine', 'euclidean', 'euclidean_l2']
|
||||||
|
|
||||||
for detector in detectors:
|
dataset = [
|
||||||
img = DeepFace.detectFace("dataset/img11.jpg", detector_backend = detector)
|
['dataset/img1.jpg', 'dataset/img2.jpg', True],
|
||||||
print(detector," test is done")
|
['dataset/img5.jpg', 'dataset/img6.jpg', True],
|
||||||
|
['dataset/img6.jpg', 'dataset/img7.jpg', True],
|
||||||
|
['dataset/img8.jpg', 'dataset/img9.jpg', True],
|
||||||
|
['dataset/img1.jpg', 'dataset/img11.jpg', True],
|
||||||
|
['dataset/img2.jpg', 'dataset/img11.jpg', True],
|
||||||
|
|
||||||
#import matplotlib.pyplot as plt
|
['dataset/img1.jpg', 'dataset/img3.jpg', False],
|
||||||
#plt.imshow(img)
|
['dataset/img2.jpg', 'dataset/img3.jpg', False],
|
||||||
#plt.show()
|
['dataset/img6.jpg', 'dataset/img8.jpg', False],
|
||||||
|
['dataset/img6.jpg', 'dataset/img9.jpg', False],
|
||||||
|
]
|
||||||
|
|
||||||
#-----------------------------------------
|
print("-----------------------------------------")
|
||||||
print("-----------------------------------------")
|
|
||||||
|
|
||||||
img_path = "dataset/img1.jpg"
|
def test_cases():
|
||||||
embedding = DeepFace.represent(img_path)
|
|
||||||
print("Function returned ", len(embedding), "dimensional vector")
|
|
||||||
|
|
||||||
model_name = "VGG-Face"
|
print("DeepFace.detectFace test")
|
||||||
model = DeepFace.build_model(model_name)
|
|
||||||
print(model_name," is built")
|
|
||||||
embedding = DeepFace.represent(img_path, model = model)
|
|
||||||
print("Represent function returned ", len(embedding), "dimensional vector")
|
|
||||||
|
|
||||||
#-----------------------------------------
|
for detector in detectors:
|
||||||
|
img = DeepFace.detectFace("dataset/img11.jpg", detector_backend = detector)
|
||||||
|
evaluate(img.shape[0] > 0 and img.shape[1] > 0)
|
||||||
|
print(detector," test is done")
|
||||||
|
|
||||||
dataset = [
|
print("-----------------------------------------")
|
||||||
['dataset/img1.jpg', 'dataset/img2.jpg', True],
|
|
||||||
['dataset/img1.jpg', 'dataset/img6.jpg', True]
|
|
||||||
]
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
img_path = "dataset/img1.jpg"
|
||||||
|
embedding = DeepFace.represent(img_path)
|
||||||
|
print("Function returned ", len(embedding), "dimensional vector")
|
||||||
|
evaluate(len(embedding) > 0)
|
||||||
|
|
||||||
print("Face detectors test")
|
print("-----------------------------------------")
|
||||||
|
|
||||||
print("retinaface detector")
|
print("Face detectors test")
|
||||||
res = DeepFace.verify(dataset, detector_backend = 'retinaface')
|
|
||||||
print(res)
|
|
||||||
|
|
||||||
print("ssd detector")
|
for detector in detectors:
|
||||||
res = DeepFace.verify(dataset, detector_backend = 'ssd')
|
print(detector + " detector")
|
||||||
print(res)
|
res = DeepFace.verify(dataset[0][0], dataset[0][1], detector_backend = detector)
|
||||||
|
print(res)
|
||||||
|
assert res["verified"] == dataset[0][2]
|
||||||
|
|
||||||
print("opencv detector")
|
print("-----------------------------------------")
|
||||||
res = DeepFace.verify(dataset, detector_backend = 'opencv')
|
|
||||||
print(res)
|
|
||||||
|
|
||||||
if False:
|
print("Find function test")
|
||||||
print("dlib detector")
|
|
||||||
res = DeepFace.verify(dataset, detector_backend = 'dlib')
|
|
||||||
print(res)
|
|
||||||
|
|
||||||
print("mtcnn detector")
|
df = DeepFace.find(img_path = "dataset/img1.jpg", db_path = "dataset")
|
||||||
res = DeepFace.verify(dataset, detector_backend = 'mtcnn')
|
print(df.head())
|
||||||
print(res)
|
evaluate(df.shape[0] > 0)
|
||||||
|
|
||||||
print("-----------------------------------------")
|
print("-----------------------------------------")
|
||||||
|
|
||||||
print("Single find function test")
|
print("Facial analysis test. Passing nothing as an action")
|
||||||
|
|
||||||
df = DeepFace.find(img_path = "dataset/img1.jpg", db_path = "dataset"
|
img = "dataset/img4.jpg"
|
||||||
#, model_name = 'Dlib'
|
demography = DeepFace.analyze(img)
|
||||||
)
|
print(demography)
|
||||||
print(df.head())
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
evaluate(demography["age"] > 20 and demography["age"] < 40)
|
||||||
|
evaluate(demography["gender"] == "Woman")
|
||||||
|
|
||||||
print("Pre-built model for single find function test")
|
print("-----------------------------------------")
|
||||||
|
|
||||||
#model_name = "VGG-Face"
|
print("Facial analysis test. Passing all to the action")
|
||||||
#model = DeepFace.build_model(model_name)
|
demography = DeepFace.analyze(img, ['age', 'gender', 'race', 'emotion'])
|
||||||
#print(model_name," is built")
|
|
||||||
|
|
||||||
df = DeepFace.find(img_path = "dataset/img1.jpg", db_path = "dataset"
|
print("Demography:")
|
||||||
, model_name = model_name, model = model
|
print(demography)
|
||||||
)
|
|
||||||
print(df.head())
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
#check response is a valid json
|
||||||
|
print("Age: ", demography["age"])
|
||||||
|
print("Gender: ", demography["gender"])
|
||||||
|
print("Race: ", demography["dominant_race"])
|
||||||
|
print("Emotion: ", demography["dominant_emotion"])
|
||||||
|
|
||||||
print("Bulk find function tests")
|
evaluate(demography.get("age") is not None)
|
||||||
|
evaluate(demography.get("gender") is not None)
|
||||||
|
evaluate(demography.get("dominant_race") is not None)
|
||||||
|
evaluate(demography.get("dominant_emotion") is not None)
|
||||||
|
|
||||||
dfs = DeepFace.find(img_path = ["dataset/img1.jpg", "dataset/img2.jpg"], db_path = "dataset"
|
print("-----------------------------------------")
|
||||||
#, model_name = 'Dlib'
|
|
||||||
)
|
|
||||||
print(dfs[0].head())
|
|
||||||
print(dfs[1].head())
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
print("Facial analysis test 2. Remove some actions and check they are not computed")
|
||||||
|
demography = DeepFace.analyze(img, ['age', 'gender'])
|
||||||
|
|
||||||
print("Bulk verification tests")
|
print("Age: ", demography.get("age"))
|
||||||
|
print("Gender: ", demography.get("gender"))
|
||||||
|
print("Race: ", demography.get("dominant_race"))
|
||||||
|
print("Emotion: ", demography.get("dominant_emotion"))
|
||||||
|
|
||||||
resp_obj = DeepFace.verify(dataset)
|
evaluate(demography.get("age") is not None)
|
||||||
print(resp_obj)
|
evaluate(demography.get("gender") is not None)
|
||||||
print(resp_obj["pair_1"]["verified"] == True)
|
evaluate(demography.get("dominant_race") is None)
|
||||||
print(resp_obj["pair_2"]["verified"] == True)
|
evaluate(demography.get("dominant_emotion") is None)
|
||||||
|
|
||||||
print("-----------------------------------------")
|
print("-----------------------------------------")
|
||||||
|
|
||||||
print("Bulk facial analysis tests")
|
print("Facial recognition tests")
|
||||||
|
|
||||||
dataset = [
|
for model in models:
|
||||||
'dataset/img1.jpg',
|
for metric in metrics:
|
||||||
'dataset/img2.jpg',
|
for instance in dataset:
|
||||||
'dataset/img5.jpg',
|
img1 = instance[0]
|
||||||
'dataset/img6.jpg'
|
img2 = instance[1]
|
||||||
]
|
result = instance[2]
|
||||||
|
|
||||||
resp_obj = DeepFace.analyze(dataset)
|
resp_obj = DeepFace.verify(img1, img2
|
||||||
print(resp_obj["instance_1"]["age"]," years old ", resp_obj["instance_1"]["dominant_emotion"], " ",resp_obj["instance_1"]["gender"])
|
, model_name = model
|
||||||
print(resp_obj["instance_2"]["age"]," years old ", resp_obj["instance_2"]["dominant_emotion"], " ",resp_obj["instance_2"]["gender"])
|
, distance_metric = metric)
|
||||||
print(resp_obj["instance_3"]["age"]," years old ", resp_obj["instance_3"]["dominant_emotion"], " ",resp_obj["instance_3"]["gender"])
|
|
||||||
print(resp_obj["instance_4"]["age"]," years old ", resp_obj["instance_4"]["dominant_emotion"], " ",resp_obj["instance_4"]["gender"])
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
prediction = resp_obj["verified"]
|
||||||
|
distance = round(resp_obj["distance"], 2)
|
||||||
|
threshold = resp_obj["threshold"]
|
||||||
|
|
||||||
print("Facial analysis test. Passing nothing as an action")
|
passed = prediction == result
|
||||||
|
|
||||||
img = "dataset/img4.jpg"
|
evaluate(passed)
|
||||||
demography = DeepFace.analyze(img)
|
|
||||||
print(demography)
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
if passed:
|
||||||
|
test_result_label = "passed"
|
||||||
|
else:
|
||||||
|
test_result_label = "failed"
|
||||||
|
|
||||||
print("Facial analysis test. Passing all to the action")
|
if prediction == True:
|
||||||
demography = DeepFace.analyze(img, ['age', 'gender', 'race', 'emotion'])
|
classified_label = "verified"
|
||||||
|
else:
|
||||||
|
classified_label = "unverified"
|
||||||
|
|
||||||
print("Demography:")
|
print(img1.split("/")[-1], "-", img2.split("/")[-1], classified_label, "as same person based on", model,"and",metric,". Distance:",distance,", Threshold:", threshold,"(",test_result_label,")")
|
||||||
print(demography)
|
|
||||||
|
|
||||||
#check response is a valid json
|
print("--------------------------")
|
||||||
print("Age: ", demography["age"])
|
|
||||||
print("Gender: ", demography["gender"])
|
|
||||||
print("Race: ", demography["dominant_race"])
|
|
||||||
print("Emotion: ", demography["dominant_emotion"])
|
|
||||||
|
|
||||||
print("-----------------------------------------")
|
# -----------------------------------------
|
||||||
|
|
||||||
|
print("Passing numpy array to analyze function")
|
||||||
|
|
||||||
print("Facial analysis test 2. Remove some actions and check they are not computed")
|
img = cv2.imread("dataset/img1.jpg")
|
||||||
demography = DeepFace.analyze(img, ['age', 'gender'])
|
resp_obj = DeepFace.analyze(img)
|
||||||
|
print(resp_obj)
|
||||||
|
|
||||||
print("Age: ", demography.get("age"))
|
evaluate(resp_obj["age"] > 20 and resp_obj["age"] < 40)
|
||||||
print("Gender: ", demography.get("gender"))
|
evaluate(resp_obj["gender"] == "Woman")
|
||||||
print("Race: ", demography.get("dominant_race"))
|
|
||||||
print("Emotion: ", demography.get("dominant_emotion"))
|
|
||||||
|
|
||||||
|
print("--------------------------")
|
||||||
|
|
||||||
print("-----------------------------------------")
|
print("Passing numpy array to verify function")
|
||||||
|
|
||||||
print("Face recognition tests")
|
img1 = cv2.imread("dataset/img1.jpg")
|
||||||
|
img2 = cv2.imread("dataset/img2.jpg")
|
||||||
|
|
||||||
dataset = [
|
res = DeepFace.verify(img1, img2)
|
||||||
['dataset/img1.jpg', 'dataset/img2.jpg', True],
|
print(res)
|
||||||
['dataset/img5.jpg', 'dataset/img6.jpg', True],
|
|
||||||
['dataset/img6.jpg', 'dataset/img7.jpg', True],
|
|
||||||
['dataset/img8.jpg', 'dataset/img9.jpg', True],
|
|
||||||
['dataset/img1.jpg', 'dataset/img11.jpg', True],
|
|
||||||
['dataset/img2.jpg', 'dataset/img11.jpg', True],
|
|
||||||
|
|
||||||
['dataset/img1.jpg', 'dataset/img3.jpg', False],
|
evaluate(res["verified"] == True)
|
||||||
['dataset/img2.jpg', 'dataset/img3.jpg', False],
|
|
||||||
['dataset/img6.jpg', 'dataset/img8.jpg', False],
|
|
||||||
['dataset/img6.jpg', 'dataset/img9.jpg', False],
|
|
||||||
]
|
|
||||||
|
|
||||||
#models = ['VGG-Face', 'Facenet', 'OpenFace', 'DeepFace', 'DeepID', 'Dlib', 'ArcFace']
|
print("--------------------------")
|
||||||
models = ['VGG-Face', 'Facenet', 'Facenet512', 'ArcFace', 'SFace'] #those are robust models
|
|
||||||
metrics = ['cosine', 'euclidean', 'euclidean_l2']
|
|
||||||
|
|
||||||
passed_tests = 0; test_cases = 0
|
print("Passing numpy array to find function")
|
||||||
|
|
||||||
for model in models:
|
img1 = cv2.imread("dataset/img1.jpg")
|
||||||
#prebuilt_model = DeepFace.build_model(model)
|
|
||||||
#print(model," is built")
|
|
||||||
for metric in metrics:
|
|
||||||
for instance in dataset:
|
|
||||||
img1 = instance[0]
|
|
||||||
img2 = instance[1]
|
|
||||||
result = instance[2]
|
|
||||||
|
|
||||||
resp_obj = DeepFace.verify(img1, img2
|
df = DeepFace.find(img1, db_path = "dataset")
|
||||||
, model_name = model
|
|
||||||
#, model = prebuilt_model
|
|
||||||
, distance_metric = metric)
|
|
||||||
|
|
||||||
prediction = resp_obj["verified"]
|
print(df.head())
|
||||||
distance = round(resp_obj["distance"], 2)
|
|
||||||
threshold = resp_obj["threshold"]
|
|
||||||
|
|
||||||
test_result_label = "failed"
|
evaluate(df.shape[0] > 0)
|
||||||
if prediction == result:
|
|
||||||
passed_tests = passed_tests + 1
|
|
||||||
test_result_label = "passed"
|
|
||||||
|
|
||||||
if prediction == True:
|
print("--------------------------")
|
||||||
classified_label = "verified"
|
|
||||||
else:
|
|
||||||
classified_label = "unverified"
|
|
||||||
|
|
||||||
test_cases = test_cases + 1
|
test_cases()
|
||||||
|
|
||||||
print(img1.split("/")[-1], "-", img2.split("/")[-1], classified_label, "as same person based on", model,"and",metric,". Distance:",distance,", Threshold:", threshold,"(",test_result_label,")")
|
print("num of test cases run: " + str(num_cases))
|
||||||
|
print("succeeded test cases: " + str(succeed_cases))
|
||||||
|
|
||||||
print("--------------------------")
|
test_score = (100 * succeed_cases) / num_cases
|
||||||
|
|
||||||
#-----------------------------------------
|
print("test coverage: " + str(test_score))
|
||||||
|
|
||||||
print("Passed unit tests: ",passed_tests," / ",test_cases)
|
if test_score > expected_coverage:
|
||||||
|
print("well done! min required test coverage is satisfied")
|
||||||
|
else:
|
||||||
|
print("min required test coverage is NOT satisfied")
|
||||||
|
|
||||||
min_score = 70
|
assert test_score > expected_coverage
|
||||||
|
|
||||||
accuracy = 100 * passed_tests / test_cases
|
|
||||||
accuracy = round(accuracy, 2)
|
|
||||||
|
|
||||||
if accuracy >= min_score:
|
|
||||||
print("Unit tests are completed successfully. Score: ",accuracy,"%")
|
|
||||||
else:
|
|
||||||
raise ValueError("Unit test score does not satisfy the minimum required accuracy. Minimum expected score is ", min_score,"% but this got ",accuracy,"%")
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
#-----------------------------------
|
|
||||||
|
|
||||||
print("Analyze function with passing pre-trained model")
|
|
||||||
|
|
||||||
emotion_model = DeepFace.build_model("Emotion")
|
|
||||||
age_model = DeepFace.build_model("Age")
|
|
||||||
gender_model = DeepFace.build_model("Gender")
|
|
||||||
race_model = DeepFace.build_model("Race")
|
|
||||||
|
|
||||||
facial_attribute_models = {}
|
|
||||||
facial_attribute_models["emotion"] = emotion_model
|
|
||||||
facial_attribute_models["age"] = age_model
|
|
||||||
facial_attribute_models["gender"] = gender_model
|
|
||||||
facial_attribute_models["race"] = race_model
|
|
||||||
|
|
||||||
resp_obj = DeepFace.analyze("dataset/img1.jpg", models=facial_attribute_models)
|
|
||||||
print(resp_obj)
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
if False:
|
|
||||||
print("Ensemble for find function")
|
|
||||||
df = DeepFace.find(img_path = "dataset/img1.jpg", db_path = "dataset", model_name = "Ensemble")
|
|
||||||
print(df.head())
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
if False:
|
|
||||||
print("Ensemble for verify function")
|
|
||||||
resp_obj = DeepFace.verify(dataset, model_name = "Ensemble")
|
|
||||||
|
|
||||||
for i in range(0, len(dataset)):
|
|
||||||
item = resp_obj['pair_%s' % (i+1)]
|
|
||||||
verified = item["verified"]
|
|
||||||
score = item["score"]
|
|
||||||
print(verified)
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
if False:
|
|
||||||
|
|
||||||
print("Pre-trained ensemble method - find")
|
|
||||||
|
|
||||||
from deepface import DeepFace
|
|
||||||
from deepface.basemodels import Boosting
|
|
||||||
|
|
||||||
model = Boosting.loadModel()
|
|
||||||
df = DeepFace.find("dataset/img1.jpg", db_path = "dataset", model_name = 'Ensemble', model = model, enforce_detection=False)
|
|
||||||
|
|
||||||
print(df)
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
if False:
|
|
||||||
print("Pre-trained ensemble method - verify")
|
|
||||||
res = DeepFace.verify(dataset, model_name = "Ensemble", model = model)
|
|
||||||
print(res)
|
|
||||||
|
|
||||||
#-----------------------------------
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
import cv2
|
|
||||||
|
|
||||||
print("Passing numpy array to analyze function")
|
|
||||||
|
|
||||||
img = cv2.imread("dataset/img1.jpg")
|
|
||||||
resp_obj = DeepFace.analyze(img)
|
|
||||||
print(resp_obj)
|
|
||||||
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
print("Passing numpy array to verify function")
|
|
||||||
|
|
||||||
img1 = cv2.imread("dataset/img1.jpg")
|
|
||||||
img2 = cv2.imread("dataset/img2.jpg")
|
|
||||||
|
|
||||||
res = DeepFace.verify(img1, img2)
|
|
||||||
print(res)
|
|
||||||
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
print("Passing numpy array to find function")
|
|
||||||
|
|
||||||
img1 = cv2.imread("dataset/img1.jpg")
|
|
||||||
|
|
||||||
df = DeepFace.find(img1, db_path = "dataset")
|
|
||||||
|
|
||||||
print(df.head())
|
|
||||||
|
|
||||||
print("--------------------------")
|
|
||||||
|
|
||||||
self.assertEqual(accuracy >= min_score, True, "A problem on the deepface installation.")
|
|
||||||
|
|
||||||
unittest.main(exit=False)
|
|
||||||
Loading…
x
Reference in New Issue
Block a user