def __init__(self):
super().__init__()
self.ui = Ui_AIOptimization()
self.ui.setupUi(self)
self.ui.tab_widget.currentChanged.connect(self._on_changed_current_tab)
LearningModel.default().training_finished.connect(
self._on_finished_training)
LearningModel.default().test_finished.connect(self._on_finished_test)
def test_object_instance(self):
for i in range(1, 100):
w = np.random.rand(1, 2) + 1
d = random.randint(-5, 5)
learning_model = LearningModel(w, d)
x = np.random.rand(1, 2)
value = learning_model.decide(x)
print(f'value {value}')
def generate_models(num_models: int,
specie='a',
generation=1) -> List[LearningModel]:
to_return = list()
size = (1, 3)
for i in range(0, num_models):
if specie == 'a':
w_vector = list(
i for i in (np.random.uniform(-2, 2, size)).flatten())
d = random.randrange(-3, 3)
model = LearningModel(w_vector, d)
model.generation = generation
to_return.append(model)
elif specie == 'b':
w_vector = list(
i for i in (np.random.uniform(-5, 5, size)).flatten())
d = random.randrange(-1, 1)
model = LearningModel(w_vector, d)
model.generation = generation
to_return.append(model)
elif specie == 'c_rare':
w_vector = list(
i for i in (np.random.uniform(-10, 10, size)).flatten())
d = random.randrange(-10, 10)
model = LearningModel(w_vector, d)
model.generation = generation
to_return.append(model)
return to_return
def __init__(self):
super().__init__()
self.ui = Ui_Dataset()
self.ui.setupUi(self)
self.all_thumbnails = []
self.selected_thumbnails: Set[Thumbnail] = set()
self.ui.image_list_widget.itemSelectionChanged.connect(
self.on_changed_image_list_selection)
self.ui.delete_images_button.clicked.connect(
self.on_clicked_delete_images_button)
self.ui.train_button.clicked.connect(self.on_clicked_train_button)
self.ui.camera_and_images_menu = QMenu()
self.ui.camera_and_images_menu.addAction(self.ui.select_images_action)
self.ui.camera_and_images_menu.addAction(self.ui.camera_action)
self.ui.camera_and_images_button.setMenu(
self.ui.camera_and_images_menu)
self.ui.select_images_action.triggered.connect(
self.on_clicked_select_images_button)
self.ui.camera_action.triggered.connect(self.on_clicked_camera_button)
self.ui.image_list_widget.setCurrentItem(
self.ui.image_list_widget.topLevelItem(0).child(
0)) # FIXME: refactor
self.ui.image_list_widget.expandAll()
self._reload_images(Dataset.Category.TRAINING_OK)
self.__reload_recent_training_date()
self.capture_dialog: Optional[ImageCaptureDialog] = None
self.preview_window = PreviewWindow()
self.watcher = QFileSystemWatcher(self)
self.watcher.addPaths([
str(Dataset.images_path(Dataset.Category.TRAINING_OK)),
str(Dataset.images_path(Dataset.Category.TEST_OK)),
str(Dataset.images_path(Dataset.Category.TEST_NG))
])
self.watcher.directoryChanged.connect(
self.on_dataset_directory_changed)
self.select_area_dialog = None
self.msgBox = None
LearningModel.default().training_finished.connect(
self.on_finished_training)
def from_bson_to_model(bson):
model_id = bson['model_id']
processed = bson['processed']
w_vector = bson['w_vector']
d = bson['d']
if processed:
score = bson['score']
else:
score = None
generation = bson['generation']
model = LearningModel(w_vector, d, model_id)
model.generation = generation
model.processed = processed
model.score = score
return model
def on_finished_selecting_area(self, data: TrimmingData):
categories = [
Dataset.Category.TRAINING_OK, Dataset.Category.TEST_OK,
Dataset.Category.TEST_NG
]
truncated_image_paths = []
for category in categories:
dir_path = Dataset.images_path(category)
save_path = Dataset.trimmed_path(category)
if os.path.exists(save_path):
shutil.rmtree(save_path)
os.mkdir(save_path)
if not data.needs_trimming:
copy_tree(str(dir_path), str(save_path))
else:
file_list = os.listdir(dir_path)
file_list = [
img for img in file_list if Path(img).suffix in
['.jpg', '.jpeg', '.png', '.gif', '.bmp']
]
for file_name in file_list:
truncated_image_path = Dataset.trim_image(
os.path.join(dir_path, file_name), save_path, data)
if truncated_image_path:
file_name = os.path.basename(truncated_image_path)
shutil.move(
truncated_image_path,
os.path.join(
Dataset.images_path(
Dataset.Category.TRUNCATED), file_name))
truncated_image_paths.append(truncated_image_path)
Project.save_latest_trimming_data(data)
# alert for moving truncated images
if truncated_image_paths:
self.msgBox = QMessageBox()
self.msgBox.setText(str(len(truncated_image_paths))+'枚の画像を読み込めませんでした. これらの画像はtruncatedフォルダに移動されました.\n\n'\
+ 'このままトレーニングを開始しますか？')
self.msgBox.setStandardButtons(self.msgBox.Yes | self.msgBox.No)
self.msgBox.setDefaultButton(self.msgBox.Yes)
reply = self.msgBox.exec()
if reply == self.msgBox.No:
return
# start training
LearningModel.default().start_training()
def setup_tool_bar(self):
self.inspection_widget_id = self.ui.main_stacked_widget.addWidget(InspectionWidget())
self.ai_optimization_widget_id = self.ui.main_stacked_widget.addWidget(AIOptimizationWidget())
self.past_result_widget_id = self.ui.main_stacked_widget.addWidget(PastResultWidget())
self.ui.optimization_action.triggered.connect(self.on_clicked_optimization_button)
self.ui.inspection_action.triggered.connect(self.on_clicked_inspection_button)
self.ui.past_result_action.triggered.connect(self.on_clicked_past_result_button)
self.ui.action_group = QActionGroup(self)
self.ui.action_group.addAction(self.ui.optimization_action)
self.ui.action_group.addAction(self.ui.inspection_action)
self.ui.action_group.addAction(self.ui.past_result_action)
self.ui.inspection_action.setChecked(True)
self.ui.action_group.setExclusive(True)
try:
self.on_clicked_inspection_button()
self.ui.inspection_action.setChecked(True)
LearningModel.default().load_weights()
except FileNotFoundError:
self.on_clicked_optimization_button()
self.ui.optimization_action.setChecked(True)
appctxt = ApplicationContext()
loader_gif_path = appctxt.get_resource('images/loader.gif')
self.loader = QMovie(loader_gif_path)
self.loader.start()
self.loader_label = QLabel()
self.loader_label.setMovie(self.loader)
self.loader_label.hide()
self.training_message = QLabel()
spacer = QWidget()
spacer.setFixedWidth(2)
self.statusBar().addPermanentWidget(self.training_message)
self.statusBar().addPermanentWidget(self.loader_label)
self.statusBar().addPermanentWidget(spacer)
self.statusBar().setSizeGripEnabled(False)
def __init__(self):
QWidget.__init__(self)
self.ui = Ui_inspection()
self.ui.setupUi(self)
# camera preview settings
self.camera_model = CameraModel.default()
self.camera_model.set_selected_camera_to_view_finder(
self.ui.camera_preview)
self.camera_model.image_saved.connect(self.on_image_saved)
self.ui.camera_preview.setFixedSize(self.__VIEW_FINDER)
self.learning_model = LearningModel.default()
self.learning_model.predicting_finished.connect(
self.on_finished_predicting)
self.select_camera_widget = CameraList()
self.select_camera_widget.clicked.connect(self.on_clicked_camera_list)
self.select_camera_widget.closed.connect(self.on_closed_camera_list)
self.ui.select_camera_button.clicked.connect(
self.on_clicked_select_camera_button)
self.ui.inspect_button.clicked.connect(self.on_clicked_inspect_button)
self.ui.inspect_existing_image_button.clicked.connect(
self.on_clicked_inspection_existing_image_button)
self.ui.result.setCurrentWidget(self.ui.default_result)
appctxt = ApplicationContext()
loader_gif_path = appctxt.get_resource('images/loader.gif')
self.loader_movie = QMovie(loader_gif_path)
self.loader_movie.setScaledSize(QSize(30, 8))
self.loader_movie.start()
self.__ng_counter = 0
self.__ok_counter = 0
self.ui.OK_counter_label.setText(str(self.ok_counter))
self.ui.NG_counter_label.setText(str(self.ng_counter))
self.msgBox = None
def __init__(self):
super().__init__()
# Disable maximizing window
self.setWindowFlags(Qt.WindowCloseButtonHint | Qt.WindowMinimizeButtonHint | Qt.CustomizeWindowHint)
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.inspection_widget_id = None
self.ai_optimization_widget_id = None
self.past_result_widget_id = None
self.msgBox = None
self.inspection_mainwindow_size = QSize(780, 550)
self.optimization_mainwindow_size = QSize(864, 730)
self.past_result_mainwindow_size = QSize(780, 600)
self.inspection_widget_size = QSize(740, 420)
self.optimization_widget_size = QSize(840, 600)
self.past_result_widget_size = QSize(740, 400)
self.latest_cpos = None
self.move_window_position()
self.setup_tool_bar()
self.setup_menu_bar()
# 一旦レポート機能なし
self.ui.past_result_action.setEnabled(False)
self.ui.past_result_action.setVisible(False)
LearningModel.default().predicting_start.connect(self.on_start_predicting)
LearningModel.default().predicting_finished.connect(self.on_finished_predicting)
LearningModel.default().training_start.connect(self.on_start_training)
LearningModel.default().training_finished.connect(self.on_finished_training)
def test_model(model: LearningModel, enemies_rectangle: List[Rectangle],
img_shape):
if len(enemies_rectangle) == 0:
return 0
p_bot_left_img = (0, img_shape[0])
enemies_rectangle.sort(key=lambda x: x.x1)
enemy_rectangle = enemies_rectangle[0]
p_top_left = (enemy_rectangle.x1, enemy_rectangle.y1)
p_top_right = (enemy_rectangle.x2, enemy_rectangle.y1)
p_bot_left = (enemy_rectangle.x1, enemy_rectangle.y2)
total_distance = distance.euclidean((0, 0), (img_shape[0], img_shape[1]))
x1 = distance.euclidean(p_bot_left_img, p_top_left) / total_distance
x2 = distance.euclidean(p_bot_left_img, p_top_right) / total_distance
x3 = (_get_angle(p_bot_left_img, p_bot_left) / 90) + 1
x_vector = np.array([[x1, x2, x3]])
to_return = model.apply(x_vector)
return to_return
def __init__(self):
self.learning_model = LearningModel.default() # initの引数で渡したい…
super().__init__()
self.ui = Ui_Test()
self.ui.setupUi(self)
appctxt = ApplicationContext()
loader_gif_path = appctxt.get_resource('images/loader.gif')
self.loader = QMovie(loader_gif_path)
self.loader.setScaledSize(QSize(30, 8))
self.ui.loading_gif_label.setMovie(self.loader)
self.loader.start()
self.ui.about_threshold_button.clicked.connect(
self.on_clicked_about_threshold_button)
self.ui.threshold_slider.valueChanged.connect(
self.on_threshold_changed)
self.ui.details_button.clicked.connect(self.on_clicked_details_button)
self.distance_figure: Figure = plt.figure(figsize=(4, 3))
self.distance_canvas = FigureCanvas(self.distance_figure)
self.distance_canvas.setParent(self.ui.distance_chart_widget)
# sns.set_palette(['#3FDA68', '#E66643'])
performance_figure = Figure(figsize=(3.5, 3.5))
self.performance_axes: Axes = performance_figure.add_subplot(111)
performance_figure.patch.set_alpha(0)
self.performance_axes.set_position(
pos=[-0.1, 0.1, 1, 1]) # FIXME: adjust position automatically
self.performance_center_circle = plt.Circle(xy=(0, 0),
radius=0.75,
fc='#F5F5F5',
linewidth=1.25)
self.performance_canvas = FigureCanvas(performance_figure)
self.performance_canvas.setParent(self.ui.performance_chart_widget)
self.test_report_widget = TestReportWidget()
def _on_changed_current_tab(self):
if self.ui.tab_widget.currentIndex() == 1: # FIXME: refactor condition
self.ui.test_tab.show_loading()
LearningModel.default().test_if_needed(predict_training=True)
def create_model(w_array, d):
w_array_mut = [i for i in w_array.ravel()]
model = LearningModel(w_array_mut, d)
model.processed = False
return model
def __init__(self):
appctxt = ApplicationContext()
env = Environment(
loader=FileSystemLoader(appctxt.get_resource('html/')))
self.template = env.get_template('test_report_template.html')
self.test_results = LearningModel.default().test_results