def __init__(self, g_pool, show_segmentation=True):
super().__init__(g_pool)
self.storage = model.Classified_Segment_Storage(plugin=self,
rec_dir=g_pool.rec_dir)
self.seek_controller = controller.Eye_Movement_Seek_Controller(
plugin=self,
storage=self.storage,
seek_to_timestamp=self.seek_to_timestamp)
self.offline_controller = controller.Eye_Movement_Offline_Controller(
plugin=self,
storage=self.storage,
on_status=self.on_task_status,
on_progress=self.on_task_progress,
on_exception=self.on_task_exception,
on_completed=self.on_task_completed,
)
self.menu_content = ui.Menu_Content(
plugin=self,
label_text=self.MENU_LABEL_TEXT,
show_segmentation=show_segmentation,
)
self.prev_segment_button = ui.Prev_Segment_Button(
on_click=self.seek_controller.jump_to_prev_segment)
self.next_segment_button = ui.Next_Segment_Button(
on_click=self.seek_controller.jump_to_next_segment)
self._gaze_changed_listener = Listener(plugin=self,
topic="gaze_positions",
rec_dir=g_pool.rec_dir)
self._gaze_changed_listener.add_observer(
method_name="on_data_changed",
observer=self.offline_controller.classify)
self._eye_movement_changed_announcer = Announcer(
plugin=self,
topic=EYE_MOVEMENT_ANNOUNCER_TOPIC,
rec_dir=g_pool.rec_dir)
class Offline_Eye_Movement_Detector(Observable, Eye_Movement_Detector_Base):
"""
Eye movement classification detector based on segmented linear regression.
Event identification is based on segmentation that simultaneously denoises the signal and determines event
boundaries. The full gaze position time-series is segmented into an approximately optimal piecewise linear
function in O(n) time. Gaze feature parameters for classification into fixations, saccades, smooth pursuits and post-saccadic oscillations
are derived from human labeling in a data-driven manner.
More details about this approach can be found here:
https://www.nature.com/articles/s41598-017-17983-x
The open source implementation can be found here:
https://gitlab.com/nslr/nslr-hmm
"""
MENU_LABEL_TEXT = "Eye Movement Detector"
def __init__(self, g_pool, show_segmentation=True):
super().__init__(g_pool)
self.storage = model.Classified_Segment_Storage(plugin=self,
rec_dir=g_pool.rec_dir)
self.add_observer("on_task_completed", self.storage.save_to_disk)
self.seek_controller = controller.Eye_Movement_Seek_Controller(
plugin=self,
storage=self.storage,
seek_to_timestamp=self.seek_to_timestamp)
self.offline_controller = controller.Eye_Movement_Offline_Controller(
plugin=self,
storage=self.storage,
on_started=self.on_task_started,
on_status=self.on_task_status,
on_progress=self.on_task_progress,
on_exception=self.on_task_exception,
on_completed=self.on_task_completed,
)
self.menu_content = ui.Menu_Content(
plugin=self,
label_text=self.MENU_LABEL_TEXT,
show_segmentation=show_segmentation,
)
self.prev_segment_button = ui.Prev_Segment_Button(
on_click=self.seek_controller.jump_to_prev_segment)
self.next_segment_button = ui.Next_Segment_Button(
on_click=self.seek_controller.jump_to_next_segment)
self._gaze_changed_listener = Listener(plugin=self,
topic="gaze_positions",
rec_dir=g_pool.rec_dir)
self._gaze_changed_listener.add_observer(
method_name="on_data_changed",
observer=self.offline_controller.classify)
self._eye_movement_changed_announcer = Announcer(
plugin=self,
topic=EYE_MOVEMENT_ANNOUNCER_TOPIC,
rec_dir=g_pool.rec_dir)
#
def trigger_recalculate(self):
self.notify_all({
"subject": Notification_Subject.SHOULD_RECALCULATE,
"delay": 0.5
})
def seek_to_timestamp(self, timestamp):
self.notify_all({
"subject": "seek_control.should_seek",
"timestamp": timestamp
})
def on_task_started(self):
self.menu_content.update_error_text("")
def on_task_progress(self, progress: float):
self.menu_content.update_progress(progress)
def on_task_status(self, status: str):
self.menu_content.update_status(status)
def on_task_exception(self, exception: Exception):
error_message = f"{exception}"
logger.error(error_message)
self.menu_content.update_error_text(error_message)
def on_task_completed(self):
self._eye_movement_changed_announcer.announce_new()
#
def init_ui(self):
self.add_menu()
self.menu_content.add_to_menu(self.menu)
self.prev_segment_button.add_to_quickbar(self.g_pool.quickbar)
self.next_segment_button.add_to_quickbar(self.g_pool.quickbar)
if len(self.storage):
status = "Loaded from cache"
self.menu_content.update_status(status)
else:
self.trigger_recalculate()
def deinit_ui(self):
self.remove_menu()
self.prev_segment_button.remove_from_quickbar(self.g_pool.quickbar)
self.next_segment_button.remove_from_quickbar(self.g_pool.quickbar)
def get_init_dict(self):
return {"show_segmentation": self.menu_content.show_segmentation}
def on_notify(self, notification):
if notification["subject"] in (
Notification_Subject.SHOULD_RECALCULATE,
Notification_Subject.MIN_DATA_CONFIDENCE_CHANGED,
):
self.offline_controller.classify()
elif notification["subject"] == "should_export":
self.export_eye_movement(notification["ts_window"],
notification["export_dir"])
def recent_events(self, events):
frame = events.get("frame")
if not frame:
return
visible_segments = self.storage.segments_in_frame(frame)
self.seek_controller.update_visible_segments(visible_segments)
self.menu_content.update_detail_text(
current_index=self.seek_controller.current_segment_index,
total_segment_count=self.seek_controller.total_segment_count,
current_segment=self.seek_controller.current_segment,
prev_segment=self.seek_controller.prev_segment,
next_segment=self.seek_controller.next_segment,
)
if self.menu_content.show_segmentation:
segment_renderer = ui.Segment_Overlay_Image_Renderer(
canvas_size=(frame.width, frame.height), image=frame.img)
for segment in visible_segments:
segment_renderer.draw(segment)
events[utils.EYE_MOVEMENT_EVENT_KEY] = visible_segments
def export_eye_movement(self, export_window, export_dir):
segments_in_section = self.storage.segments_in_timestamp_window(
export_window)
if segments_in_section:
by_segment_csv_exporter = controller.Eye_Movement_By_Segment_CSV_Exporter(
)
by_segment_csv_exporter.csv_export(segments_in_section,
export_dir=export_dir)
export_window_start, export_window_stop = export_window
ts_segment_class_pairs = ((gaze["timestamp"], seg.segment_class)
for seg in segments_in_section
for gaze in seg.segment_data
if export_window_start <=
gaze["timestamp"] <= export_window_stop)
by_gaze_csv_exporter = controller.Eye_Movement_By_Gaze_CSV_Exporter(
)
by_gaze_csv_exporter.csv_export(ts_segment_class_pairs,
export_dir=export_dir)
else:
logger.warning(
"The selected export range does not include eye movement detections"
)
def cleanup(self):
self.remove_observer("on_task_completed", self.storage.save_to_disk)