[tracks] react on move request and move window

something is still off with the deletion...

fixtracks/utils/enums.py

@@ -7,6 +7,8 @@ class DetectionData(Enum):
     FRAME = 1
     COORDINATES = 2
     TRACK_ID = 3
+    USERLABELED = 4
+    SCORE = 5
 
 class Tracks(Enum):
     TRACKONE = 1

fixtracks/utils/signals.py

@@ -22,6 +22,8 @@ class DetectionSceneSignals(QObject):
 
 class DetectionTimelineSignals(QObject):
     windowMoved = Signal()
+    manualMove = Signal()
+    moveRequest = Signal(float)
 
 class DetectionSignals(QObject):
     hover = Signal((int, QPointF))

fixtracks/utils/trackingdata.py

@@ -14,8 +14,8 @@ class TrackingData(QObject):
         if "userlabeled" not in self._data.keys():
             self._data["userlabeled"] = np.zeros_like(self["frame"], dtype=bool)
         self._columns = [k for k in self._data.keys()]
-        self._indices = self["index"]
-        self._selection = np.asarray([])
+        self._indices = np.arange(len(self["index"]), dtype=int)
+        self._selection_indices = np.asarray([])
         self._selected_ids = None
 
     @property
@@ -38,11 +38,13 @@ class TrackingData(QObject):
         return self._data["track"].shape[0]
 
     def _find(self, ids):
+        if len(ids) < 1:
+            return np.array([])
         ids = np.sort(ids)
         indexes = np.ones_like(ids, dtype=int) * -1
         j = 0
-        for idx, i in enumerate(self._indices):
-            if i == ids[j]:
+        for idx in self._indices:
+            if self["index"][idx] == ids[j]:
                 indexes[j] = idx
                 j += 1
             if j == len(indexes):
@@ -52,19 +54,23 @@ class TrackingData(QObject):
 
     @property
     def selectionIndices(self):
-        return self._selection
+        return self._selection_indices
+
+    @property
+    def selectionIDs(self):
+        return self._selected_ids
 
     def setSelectionRange(self, col, start, stop):
-        logging.debug("Trackingdata: set selection range based on column %s to %.2f - %.2f", col, start, stop)
+        logging.info("Trackingdata: set selection range based on column %s to %.2f - %.2f", col, start, stop)
         col_indices = np.where((self[col] >= start) & (self[col] < stop))[0]
-        self._selection = self._indices[col_indices]
+        self._selection_indices = self._indices[col_indices]
         if len(col_indices) < 1:
             logging.warning("TrackingData: Selection range is empty!")
 
     def selectedData(self, col:str):
         if col not in self.columns:
             logging.error("TrackingData:selectedData: Invalid column name! %s", col)
-        return self[col][self._selection]
+        return self[col][self._selection_indices]
 
     def setSelection(self, ids):
         """
@@ -75,8 +81,10 @@ class TrackingData(QObject):
         ids : array-like
             An array-like object containing the IDs to be set as user selections. 
         """
-        self._selection = self._find(ids)
+        logging.debug("TrackingData.setSelection: %i number of ids", len(ids))
+        self._selection_indices = self._find(ids)
         self._selected_ids = ids
+        # print(self._selected_ids, self._selection_indices)
 
     def setTrack(self, track_id:int, setUserLabeled:bool=True)-> None:
         """Assign a new track_id to the user-selected detections
@@ -88,9 +96,13 @@ class TrackingData(QObject):
         setUserLabeled : bool
             Should the "userlabeled" state of the detections be set to True? Otherwise they will be left untouched.
         """
-        self["track"][self._user_selections] = track_id
+        logging.info("TrackingData: set track id %i for selection, set user-labeled status %s", track_id, str(setUserLabeled))
+        # print(self._selected_ids, self._selection_indices)
+        # print("before: ", self["track"][self._selection_indices], self["frame"][self._selection_indices])
+        self["track"][self._selection_indices] = track_id
         if setUserLabeled:
             self.setUserLabeledStatus(True, True)
+        # print("after: ", self["track"][self._selection_indices], self["frame"][self._selection_indices])
 
     def setUserLabeledStatus(self, new_status: bool, selection=True):
         """Sets the status of the "userlabeled" column to a given value (True|False). This can done for ALL data in one go, or only for the UserSelection.
@@ -105,11 +117,11 @@ class TrackingData(QObject):
         logging.debug("TrackingData: (Re-)setting assignment status of %s to %s", 
                       "user selected data" if selection else " ALL", str(new_status))
         if selection:
-            self["userlabeled"][self._selection] = new_status
+            self["userlabeled"][self._selection_indices] = new_status
         else:
             self["userlabeled"][:] = new_status
 
-    def revertAssignmentStatus(self):
+    def revertUserLabeledStatus(self):
         logging.debug("TrackingData:Un-setting assignment status of all data!")
         self["userlabeled"][:] = False
 
@@ -119,12 +131,14 @@ class TrackingData(QObject):
 
     def deleteDetections(self, ids=None):
         if ids is not None:
+            logging.debug("TrackingData.deleteDetections of %i detections", len(ids))
             del_indices = self._find(ids)
         else:
-            del_indices = self._indices
+            logging.debug("TrackingData.deleteDetections of all selected detections (%i)", len(self._selected_ids))
+            del_indices = self._selected_ids
         for c in self._columns:
             self._data[c] = np.delete(self._data[c], del_indices, axis=0)
-        self._indices = self["index"]
+        self._indices = self._indices[:-len(del_indices)]
         self._selected_ids = np.setdiff1d(self._selected_ids, del_indices)
 
     def assignTracks(self, tracks:np.ndarray):
@@ -167,10 +181,10 @@ class TrackingData(QObject):
                         and M is number of keypoints
         """
         if selection:
-            if len(self._selection) < 1:
+            if len(self._selection_indices) < 1:
                 logging.info("TrackingData.coordinates returns empty array, not detections in range!")
                 return np.ndarray([])
-            return np.stack(self["keypoints"][self._selection]).astype(np.float32)
+            return np.stack(self["keypoints"][self._selection_indices]).astype(np.float32)
         return np.stack(self["keypoints"]).astype(np.float32)
 
     def keypointScores(self, selection=False):
@@ -184,10 +198,10 @@ class TrackingData(QObject):
             with N the number of detections and M the number of keypoints.
         """
         if selection:
-            if len(self._selection) < 1:
+            if len(self._selection_indices) < 1:
                 logging.info("TrackingData.scores returns empty array, not detections in range!")
                 return None
-            return np.stack(self["keypoint_score"][self._selection]).astype(np.float32)
+            return np.stack(self["keypoint_score"][self._selection_indices]).astype(np.float32)
         return np.stack(self["keypoint_score"]).astype(np.float32)
 
     def centerOfGravity(self, selection=False, threshold=0.8, nodes=[0,1,2]):

fixtracks/widgets/classifier.py

@@ -1,8 +1,8 @@
 import logging
 import numpy as np
 
-from PySide6.QtWidgets import QWidget, QVBoxLayout, QTabWidget, QPushButton, QGraphicsView
-from PySide6.QtWidgets import QSpinBox, QProgressBar, QGridLayout, QLabel, QCheckBox, QProgressDialog
+from PySide6.QtWidgets import QWidget, QVBoxLayout, QTabWidget, QPushButton, QGraphicsView, QTextEdit
+from PySide6.QtWidgets import QSpinBox, QProgressBar, QGridLayout, QLabel, QCheckBox, QDoubleSpinBox
 from PySide6.QtCore import Qt, Signal, Slot, QRunnable, QObject, QThreadPool
 from PySide6.QtGui import QBrush, QColor
 
@@ -12,23 +12,37 @@ from fixtracks.utils.trackingdata import TrackingData
 
 from IPython import embed
 
+class Detection():
+    def __init__(self, id, frame, track, position, orientation, length, userlabeled, confidence):
+        self.id = id
+        self.frame = frame
+        self.track = track
+        self.position = position
+        self.confidence = confidence
+        self.angle = orientation
+        self.length = length
+        self.userlabeled = userlabeled
+
+
 class WorkerSignals(QObject):
-    error = Signal(str)
+    message = Signal(str)
     running = Signal(bool)
     progress = Signal(int, int, int)
     currentframe = Signal(int)
     stopped = Signal(int)
 
-class ConsitencyDataLoader(QRunnable):
+
+class ConsistencyDataLoader(QRunnable):
     def __init__(self, data):
         super().__init__()
         self.signals = WorkerSignals()
         self.data = data
-        self.bendedness = self.positions = None
+        self.bendedness = None
+        self.positions = None
         self.lengths = None
         self.orientations = None
         self.userlabeled = None
-        self.scores = None
+        self.confidence = None
         self.frames = None
         self.tracks = None
 
@@ -40,17 +54,18 @@ class ConsitencyDataLoader(QRunnable):
         self.positions = self.data.centerOfGravity()
         self.orientations = self.data.orientation()
         self.lengths = self.data.animalLength()
-        self.bendedness = self.data.bendedness()
+        # self.bendedness = self.data.bendedness()
         self.userlabeled = self.data["userlabeled"]
-        self.scores = self.data["confidence"]  # ignore for now, let's see how far this carries.
+        self.confidence = self.data["confidence"]  # ignore for now, let's see how far this carries.
         self.frames = self.data["frame"]
         self.tracks = self.data["track"]
         self.signals.stopped.emit(0)
 
+
 class ConsistencyWorker(QRunnable):
 
     def __init__(self, positions, orientations, lengths, bendedness, frames, tracks,
-                 userlabeled, startframe=0, stoponerror=False) -> None:
+                 userlabeled, confidence, startframe=0, stoponerror=False, min_confidence=0.0) -> None:
         super().__init__()
         self.signals = WorkerSignals()
         self.positions = positions
@@ -58,6 +73,8 @@ class ConsistencyWorker(QRunnable):
         self.lengths = lengths
         self.bendedness = bendedness
         self.userlabeled = userlabeled
+        self.confidence = confidence
+        self._min_confidence = min_confidence
         self.frames = frames
         self.tracks = tracks
         self._startframe = startframe
@@ -70,124 +87,195 @@ class ConsistencyWorker(QRunnable):
 
     @Slot()
     def run(self):
-        def needs_checking(original, new):
-            res = False
-            for n, o in zip(new, original):
-                res = (o == 1 or o == 2) and n != o
-            if res:
-                print("inverted assignment, needs cross-checking?")
-            if not res:
-                res = len(new) > 1 and (np.all(new == 1) or np.all(new == 2))
-            if res:
-                print("all detections would be assigned to one track!")
-            return res
+        def get_detections(frame, indices):
+            detections = []
+            for i in indices:
+                if np.any(self.positions[i] < 0.1):
+                    logging.debug("Encountered probably invalid position %s", str(self.positions[i]))
+                    continue
+                if self._min_confidence > 0.0 and self.confidence[i] < self._min_confidence:
+                    self.tracks[i] = -1
+                    continue
+                d = Detection(i, frame, self.tracks[i], self.positions[i],
+                              self.orientations[i], self.lengths[i],
+                              self.userlabeled[i], self.confidence[i])
+                detections.append(d)
+            return detections
 
-        def assign_by_distance(f, p):
-            t1_step = f - last_frame[0]
-            t2_step = f - last_frame[1]
+        def assign_by_distance(d):
+            t1_step = d.frame - last_detections[1].frame
+            t2_step = d.frame - last_detections[2].frame
             if t1_step == 0 or t2_step == 0:
-                print(f"framecount is zero! current frame {f}, last frame {last_frame[0]} and {last_frame[1]}")
-
-            distance_to_trackone = np.linalg.norm(p - last_pos[0])/t1_step
-            distance_to_tracktwo = np.linalg.norm(p - last_pos[1])/t2_step
+                print(f"framecount is zero! current frame {f}, last frame {last_detections[1].frame} and {last_detections[2].frame}")
+            distance_to_trackone = np.linalg.norm(d.position - last_detections[1].position) /t1_step
+            distance_to_tracktwo = np.linalg.norm(d.position - last_detections[2].position) /t2_step
             most_likely_track = np.argmin([distance_to_trackone, distance_to_tracktwo]) + 1
             distances = np.zeros(2)
             distances[0] = distance_to_trackone
             distances[1] = distance_to_tracktwo
             return most_likely_track, distances
 
-        def assign_by_orientation(f, o):
-            t1_step = f - last_frame[0]
-            t2_step = f - last_frame[1]
-            orientationchange = (last_angle - o)
-            orientationchange[orientationchange > 180] = 360 - orientationchange[orientationchange > 180]
-            orientationchange /= np.array([t1_step, t2_step])
-            # orientationchange = np.abs(np.unwrap((last_angle - o)/np.array([t1_step, t2_step])))
-            most_likely_track = np.argmin(np.abs(orientationchange)) + 1
-            return most_likely_track, orientationchange
+        def assign_by_orientation(d):
+            t1_step = d.frame - last_detections[1].frame
+            t2_step = d.frame - last_detections[2].frame
+            orientationchanges = np.zeros(2)
+            for i in [1, 2]:
+                orientationchanges[i-1] = (last_detections[i].angle - d.angle)
 
-        def assign_by_length(o):
-            length_difference = np.abs((last_length - o))
-            most_likely_track = np.argmin(length_difference) + 1
-            return most_likely_track, length_difference
+            orientationchanges[orientationchanges > 180] = 360 - orientationchanges[orientationchanges > 180]
+            orientationchanges /= np.array([t1_step, t2_step])
+            most_likely_track = np.argmin(np.abs(orientationchanges)) + 1
+            return most_likely_track, orientationchanges
 
-        def do_assignment(f, indices, assignments):
-            for i, idx in enumerate(indices):
-                self.tracks[idx] = assignments[i]
-                last_pos[assignments[i]-1] = pp[i]
-                last_frame[assignments[i]-1] = f
-                last_angle[assignments[i]-1] = self.orientations[idx]
-                last_length[assignments[i]-1] += ((self.lengths[idx] - last_length[assignments[i]-1])/processed)
+        def assign_by_length(d):
+            length_differences = np.zeros(2)
+            length_differences[0] = np.abs((last_detections[1].length - d.length))
+            length_differences[1] = np.abs((last_detections[2].length - d.length))
+            most_likely_track = np.argmin(length_differences) + 1
+            return most_likely_track, length_differences
 
-        last_pos = [self.positions[(self.tracks == 1) & (self.frames <= self._startframe)][-1],
-                    self.positions[(self.tracks == 2) & (self.frames <= self._startframe)][-1]]
-        last_frame = [self.frames[(self.tracks == 1) & (self.frames <= self._startframe)][-1],
-                      self.frames[(self.tracks == 2) & (self.frames <= self._startframe)][-1]]
-        last_angle = [self.orientations[(self.tracks == 1) & (self.frames <= self._startframe)][-1],
-                      self.orientations[(self.tracks == 2) & (self.frames <= self._startframe)][-1]]
-        last_length = [self.lengths[(self.tracks == 1) & (self.frames <= self._startframe)][-1],
-                       self.lengths[(self.tracks == 2) & (self.frames <= self._startframe)][-1]]
+        def check_multiple_detections(detections):
+            if self._min_confidence > 0.0:
+                for i, d in enumerate(detections):
+                    if d.confidence < self._min_confidence:
+                        del detections[i]
+            distances = np.zeros((len(detections), len(detections)))
+            for i, d1 in enumerate(detections):
+                for j, d2 in enumerate(detections):
+                    distances[i, j] = np.abs(np.linalg.norm(d2.position - d1.position))
+            lowest_dist = np.argmin(np.sum(distances, axis=1))
+            del detections[lowest_dist]
+            return detections
 
+        def find_last_userlabeled(startframe):
+            t1index = np.where((self.frames < startframe) & (self.userlabeled) & (self.tracks == 1))[0][-1]
+            t2index = np.where((self.frames < startframe) & (self.userlabeled) & (self.tracks == 2))[0][-1]
+            d1 = Detection(t1index, self.frames[t1index], self.tracks[t1index], self.positions[t1index],
+                           self.orientations[t1index], self.lengths[t1index], self.userlabeled[t1index], 
+                           self.confidence[t1index])
+            d2 = Detection(t1index, self.frames[t2index], self.tracks[t2index], self.positions[t2index],
+                           self.orientations[t2index], self.lengths[t2index], self.userlabeled[t2index],
+                           self.confidence[t1index])
+            last_detections[1] = d1
+            last_detections[2] = d2
+
+        unique_frames = np.unique(self.frames)
+        steps = int((len(unique_frames) - self._startframe) // 100)
         errors = 0
         processed = 1
         progress = 0
         self._stoprequest = False
-        maxframes = np.max(self.frames)
-        startframe = np.max(last_frame)
-        steps = int((maxframes - startframe) // 200)
+        last_detections = {1: None, 2: None, -1: None}
+        find_last_userlabeled(self._startframe)
 
-        for f in np.unique(self.frames[self.frames > startframe]):
-            processed += 1
-            self.signals.currentframe.emit(f)
+        for f in unique_frames[unique_frames >= self._startframe]:
             if self._stoprequest:
                 break
+            error = False
+            message = ""
+            self.signals.currentframe.emit(f)
             indices = np.where(self.frames == f)[0]
-            pp  = self.positions[indices]
-            originaltracks = self.tracks[indices]
-            dist_assignments = np.zeros_like(originaltracks)
-            angle_assignments = np.zeros_like(originaltracks)
-            length_assignments = np.zeros_like(originaltracks)
-            userlabeled = np.zeros_like(originaltracks)
-            distances = np.zeros((len(originaltracks), 2))
+            detections = get_detections(f, indices)
+            done = [False, False]
+            if len(detections) == 0:
+                continue
+            if len(detections) > 2:
+                message = f"Frame {f}: More than 2 detections ({len(detections)}) in the same frame!"
+                logging.info("ConsistencyTracker: %s", message)
+                self.signals.message.emit(message)
+                while len(detections) > 2:
+                    detections = check_multiple_detections(detections)
+
+            if len(detections) > 1 and np.any([detections[0].userlabeled, detections[1].userlabeled]):
+                # more than one detection
+                if detections[0].userlabeled and detections[1].userlabeled:
+                    if detections[0].track == detections[1].track:
+                        error = True
+                        message = f"Frame {f}: Classification error both detections in the same frame are assigned to the same track!"
+                        logging.info("ConsistencyTracker: %s", message)
+                        self.signals.message.emit(message)
+                elif detections[0].userlabeled and not detections[1].userlabeled:
+                    detections[1].track = 1 if detections[0].track == 2 else 2
+                elif not detections[0].userlabeled and detections[1].userlabeled:
+                    detections[0].track = 1 if detections[1].track == 2 else 2
+
+                if not error:
+                    last_detections[detections[0].track] = detections[0]
+                    last_detections[detections[1].track] = detections[1]
+                    self.tracks[detections[0].id] = detections[0].track
+                    self.tracks[detections[1].id] = detections[1].track
+                    done[0] = True
+                    done[1] = True
+            elif len(detections) == 1 and detections[0].userlabeled:  # ony one detection and labeled
+                last_detections[detections[0].track] = detections[0]
+                done[0] = True
+            if np.sum(done) == len(detections):
+                continue
+
+            if error and self._stoponerror:
+                self.signals.message.emit(f"Tracking stopped at frame {f}.")
+                break
+            elif error:
+                continue
+            dist_assignments = np.zeros(2, dtype=int)
+            orientation_assignments = np.zeros_like(dist_assignments)
+            length_assignments = np.zeros_like(dist_assignments)
+            distances = np.zeros((2, 2))
             orientations = np.zeros_like(distances)
             lengths = np.zeros_like(distances)
+            assignments = np.zeros(2)
+            for i, d in enumerate(detections):
+                dist_assignments[i], distances[i, :] = assign_by_distance(d)
+                orientation_assignments[i], orientations[i,:] = assign_by_orientation(d)
+                length_assignments[i], lengths[i, :] = assign_by_length(d)
+            assignments = dist_assignments  # (dist_assignments * 10 + orientation_assignments + length_assignments) / 3
 
-            for i, (idx, p) in enumerate(zip(indices, pp)):
-                if self.userlabeled[idx]:
-                    print("user")
-                    userlabeled[i] = True
-                    last_pos[originaltracks[i]-1] = pp[i]
-                    last_frame[originaltracks[i]-1] = f
-                    last_angle[originaltracks[i]-1] = self.orientations[idx]
-                    last_length[originaltracks[i]-1] += ((self.lengths[idx] - last_length[originaltracks[i]-1]) / processed)
-                    continue
-                dist_assignments[i], distances[i, :] = assign_by_distance(f, p)
-                angle_assignments[i], orientations[i,:] = assign_by_orientation(f, self.orientations[idx])
-                length_assignments[i], lengths[i, :] = assign_by_length(self.lengths[idx])
-            if np.any(userlabeled):
-                continue
-            # check (re) assignment, update, and proceed
-            if not needs_checking(originaltracks, dist_assignments):
-                do_assignment(f, indices, dist_assignments)
-            else:
-                if not (np.all(length_assignments == 1) or np.all(length_assignments == 2)):  # if I find a solution by body length
-                    logging.debug("frame %i: Decision based on body length", f)
-                    do_assignment(f, indices, length_assignments)
-                elif not (np.all(angle_assignments == 1) or np.all(angle_assignments == 2)):  # else there is a solution based on orientation
-                    logging.info("frame %i: Decision based on orientation", f)
-                    do_assignment(f, indices, angle_assignments)
-                else:
-                    logging.info("frame %i: Cannot decide who is who")
-                    for idx in indices:
-                        self.tracks[idx] = -1
+            error = False
+            temp = {}
+            message = ""
+            if len(detections) > 1:
+                if assignments[0] == assignments[1]:
+                    d.track = -1
+                    error = True
                     errors += 1
-                    if self._stoponerror:
-                        break
+                    message = f"Frame {f}: Classification error: both detections in the same frame are assigned to the same track!"
+                    break
+                elif assignments[0] != assignments[1]:
+                    detections[0].track = assignments[0]
+                    detections[1].track = assignments[1]
+                    temp[detections[0].track] = detections[0]
+                    temp[detections[1].track] = detections[1]
+                    self.tracks[detections[0].id] = detections[0].track
+                    self.tracks[detections[1].id] = detections[1].track
+            else:
+                if np.abs(np.diff(distances[0,:])) > 50:  # maybe include the time difference into this?
+                    detections[0].track = assignments[0]
+                    temp[detections[0].track] = detections[0]
+                    self.tracks[detections[0].id] = detections[0].track
+                else:
+                    self.tracks[detections[0].id] = -1
+                    message = f"Frame: {f}: Decision based on distance not safe. Track set to unassigned."
+                    error = True
+                    errors += 1
+
+            if not error:
+                for k in temp:
+                    last_detections[temp[k].track] = temp[k]
+            else:
+                logging.info("frame %i: Cannot decide who is who! %s", f, message)
+                for idx in indices:
+                    self.tracks[idx] = -1
+                errors += 1
+                if error and self._stoponerror:
+                    self.signals.message.emit(message)
+                    break
+            processed += 1
 
             if steps > 0 and f % steps == 0:
                 progress += 1
                 self.signals.progress.emit(progress, processed, errors)
 
+        self.signals.message.emit(f"Tracking stopped at frame {f}.")
         self.signals.stopped.emit(f)
 
 
@@ -266,6 +354,7 @@ class SizeClassifier(QWidget):
         tracks[(self._sizes >= t2lower) & (self._sizes < t2upper)] = 2
         return tracks
 
+
 class NeighborhoodValidator(QWidget):
     apply = Signal()
     name = "Neighborhood Validator"
@@ -390,6 +479,7 @@ class ConsistencyClassifier(QWidget):
         self._all_lengths = None
         self._all_bendedness = None
         self._all_scores = None
+        self._confidence = None
         self._userlabeled = None
         self._maxframes = 0
         self._frames = None
@@ -435,25 +525,38 @@ class ConsistencyClassifier(QWidget):
         self._stoponerror.setChecked(True)
         self.threadpool = QThreadPool()
 
+        self._ignore_confidence = QCheckBox("Ignore detections widh confidence less than")
+        self._confidence_spinner = QDoubleSpinBox()
+        self._confidence_spinner.setRange(0.0, 1.0)
+        self._confidence_spinner.setSingleStep(0.01)
+        self._confidence_spinner.setDecimals(2)
+        self._confidence_spinner.setValue(0.5)
+        self._messagebox = QTextEdit()
+        self._messagebox.setFocusPolicy(Qt.NoFocus)
+        self._messagebox.setReadOnly(True)
+
         lyt = QGridLayout()
         lyt.addWidget(QLabel("Start frame:"), 0, 0 )
-        lyt.addWidget(self._startframe_spinner, 0, 1, 1, 2)
-        lyt.addWidget(QLabel("of"), 1, 1, 1, 1)
-        lyt.addWidget(self._maxframeslabel, 1, 2, 1, 1)
-        lyt.addWidget(self._stoponerror, 2, 0, 1, 3)
-        lyt.addWidget(QLabel("Current frame"), 3,0)
-        lyt.addWidget(self._framelabel, 3,1)
-        lyt.addWidget(QLabel("assigned"), 4, 0)
-        lyt.addWidget(self._assignedlabel, 4, 1)
-        lyt.addWidget(QLabel("errors/issues"), 5, 0)
-        lyt.addWidget(self._errorlabel, 5, 1)
+        lyt.addWidget(self._startframe_spinner, 0, 1, 1, 1)
+        lyt.addWidget(QLabel("of"), 0, 2, 1, 1)
+        lyt.addWidget(self._maxframeslabel, 0, 3, 1, 1)
+        lyt.addWidget(self._stoponerror, 1, 0, 1, 3)
+        lyt.addWidget(self._ignore_confidence, 3, 0, 1, 3)
+        lyt.addWidget(self._confidence_spinner, 3, 3, 1, 1)
+        lyt.addWidget(QLabel("Current frame"), 4, 0)
+        lyt.addWidget(self._framelabel, 4, 1)
+        lyt.addWidget(QLabel("(Re-)Assigned"), 5, 0)
+        lyt.addWidget(self._assignedlabel, 5, 1)
+        lyt.addWidget(QLabel("Errors/issues"), 5, 2)
+        lyt.addWidget(self._errorlabel, 5, 3, 1, 1)
+        lyt.addWidget(self._messagebox, 6, 0, 2, 4)
 
-        lyt.addWidget(self._startbtn, 6, 0)
-        lyt.addWidget(self._stopbtn, 6, 1)
-        lyt.addWidget(self._proceedbtn, 6, 2)
-        lyt.addWidget(self._apply_btn, 7, 0, 1, 2)
-        lyt.addWidget(self._refreshbtn, 7, 2, 1, 1)
-        lyt.addWidget(self._progressbar, 8, 0, 1, 3)
+        lyt.addWidget(self._startbtn, 8, 0, 1, 2)
+        lyt.addWidget(self._stopbtn, 8, 2)
+        # lyt.addWidget(self._proceedbtn, 8, 2)
+        lyt.addWidget(self._refreshbtn, 8, 3, 1, 1)
+        lyt.addWidget(self._apply_btn, 9, 0, 1, 4)
+        lyt.addWidget(self._progressbar, 10, 0, 1, 4)
         self.setLayout(lyt)
 
     def setData(self, data:TrackingData):
@@ -478,21 +581,38 @@ class ConsistencyClassifier(QWidget):
             self._all_lengths = self._dataworker.lengths
             self._all_bendedness = self._dataworker.bendedness
             self._userlabeled = self._dataworker.userlabeled
-            self._all_scores = self._dataworker.scores
+            self._confidence = self._dataworker.confidence
             self._frames = self._dataworker.frames
             self._tracks = self._dataworker.tracks
+            self._dataworker = None
+        if np.sum(self._userlabeled) < 1:
+            msg = "ConsistencyTracker: I need at least 1 user-labeled frame to start with!"
+            logging.error(msg)
+            self._messagebox.append(msg)
+            self.setEnabled(False)
+        else:
+            t1_userlabeled = self._frames[self._userlabeled & (self._tracks == 1)]
+            t2_userlabeled = self._frames[self._userlabeled & (self._tracks == 2)]
+            if any([len(t1_userlabeled) == 0, len(t2_userlabeled)== 0]):
+                self._messagebox.append("Error preparing data! Make sure that the first user-labeled frames contain both tracks!")
+                self.setEnabled(False)
+                return
+            max_startframe = np.min([t1_userlabeled[-1], t2_userlabeled[-1]]) -1
+            first_guess = np.max([t1_userlabeled[0], t2_userlabeled[0]])
+            while first_guess not in t1_userlabeled or first_guess not in t2_userlabeled:
+                first_guess += 1
+            min_startframe = first_guess + 1
             self._maxframes = np.max(self._frames)
-            # FIXME the following line causes an error when there are no detections in the range
-            min_frame = max([self._frames[self._tracks == 1][0], self._frames[self._tracks == 2][0]]) + 1
             self._maxframeslabel.setText(str(self._maxframes))
-            self._startframe_spinner.setMinimum(min_frame)
-            self._startframe_spinner.setMaximum(self._frames[-1])
-            self._startframe_spinner.setValue(self._frames[0] + 1)
+            self._startframe_spinner.setMinimum(min_startframe)
+            self._startframe_spinner.setMaximum(max_startframe)
+            self._startframe_spinner.setValue(min_startframe)
+            self._startframe_spinner.setSingleStep(20)
+            self._startframe_spinner.setToolTip(f"Maximum possible start frame: {max_startframe}")
             self._startbtn.setEnabled(True)
             self._assignedlabel.setText("0")
             self._errorlabel.setText("0")
-            self._dataworker = None
-        self.setEnabled(True)
+            self.setEnabled(True)
 
     @Slot(float)
     def on_progress(self, value):
@@ -502,32 +622,39 @@ class ConsistencyClassifier(QWidget):
     def stop(self):
         if self._worker is not None:
             self._worker.stop()
-        self._startbtn.setEnabled(True)
-        self._proceedbtn.setEnabled(True)
-        self._stopbtn.setEnabled(False)
-        self._refreshbtn.setEnabled(True)
+            self._messagebox.append("Stopping tracking.")
 
     def start(self):
+        confidence_level = self._confidence_spinner.value() if self._ignore_confidence.isChecked() else 0.0
         self._startbtn.setEnabled(False)
         self._refreshbtn.setEnabled(False)
         self._stopbtn.setEnabled(True)
         self._worker = ConsistencyWorker(self._all_pos, self._all_orientations, self._all_lengths,
                                          self._all_bendedness, self._frames, self._tracks, self._userlabeled,
-                                         self._startframe_spinner.value(), self._stoponerror.isChecked())
+                                         self._confidence, self._startframe_spinner.value(), self._stoponerror.isChecked(), 
+                                         min_confidence=confidence_level)
         self._worker.signals.stopped.connect(self.worker_stopped)
         self._worker.signals.progress.connect(self.worker_progress)
+        self._worker.signals.message.connect(self.worker_error)
         self._worker.signals.currentframe.connect(self.worker_frame)
+        self._messagebox.append("Tracking in progress ...")
         self.threadpool.start(self._worker)
 
     def worker_frame(self, frame):
         self._framelabel.setText(str(frame))
 
+    def worker_error(self, msg):
+        self._messagebox.append(msg)
+
     def proceed(self):
         self.start()
 
     def refresh(self):
-        self._dataworker = ConsitencyDataLoader(self._data)
+        self.setEnabled(False)
+        self._dataworker = ConsistencyDataLoader(self._data)
         self._dataworker.signals.stopped.connect(self.data_processed)
+        self._messagebox.clear()
+        self._messagebox.append("Refreshing...")
         self.threadpool.start(self._dataworker)
 
     def worker_progress(self, progress, processed, errors):
@@ -536,13 +663,15 @@ class ConsistencyClassifier(QWidget):
         self._assignedlabel.setText(str(processed))
 
     def worker_stopped(self, frame):
-        self._apply_btn.setEnabled(True)
         self._startbtn.setEnabled(True)
+        self._proceedbtn.setEnabled(True)
         self._stopbtn.setEnabled(False)
+        self._apply_btn.setEnabled(True)
+        self._refreshbtn.setEnabled(True)
         self._startframe_spinner.setValue(frame-1)
         self._proceedbtn.setEnabled(bool(frame < self._maxframes-1))
-        self._refreshbtn.setEnabled(True)
         self._processed_frames = frame
+        self._messagebox.append("... done.")
 
     def assignedTracks(self):
         return self._tracks
@@ -607,14 +736,14 @@ def main():
     import pickle
     from fixtracks.info import PACKAGE_ROOT
 
-    datafile = PACKAGE_ROOT / "data/merged2.pkl"
+    datafile = PACKAGE_ROOT / "data/merged_small_beginning.pkl"
 
     with open(datafile, "rb") as f:
         df = pickle.load(f)
-    data = TrackingData()
-    data.setData(as_dict(df))
+    data = TrackingData(as_dict(df))
     coords = data.coordinates()
-
+    cogs = data.centerOfGravity()
+    userlabeled = data["userlabeled"]
     app = QApplication([])
     window = QWidget()
     window.setMinimumSize(200, 200)
@@ -624,7 +753,7 @@ def main():
     # else:
     w = ClassifierWidget()
     w.setData(data)
-    w.size_classifier.setCoordinates(coords)
+    # w.size_classifier.setCoordinates(coords)
 
     layout = QVBoxLayout()
     layout.addWidget(w)

fixtracks/widgets/detectiontimeline.py

@@ -21,16 +21,16 @@ class Window(QGraphicsRectItem):
         self.setBrush(brush)
         self.setZValue(1.0)
         self.setAcceptHoverEvents(True)  # Enable hover events if needed
-        self.setFlags(
-            QGraphicsItem.ItemIsMovable |  # Enables item dragging
-            QGraphicsItem.ItemIsSelectable  # Enables item selection
-        )
+        # self.setFlags(
+        #     QGraphicsItem.ItemIsMovable |  # Enables item dragging
+        #     QGraphicsItem.ItemIsSelectable  # Enables item selection
+        # )
         self._y = y
 
     def setWindowX(self, newx):
         logging.debug("timeline.window: set position to %.3f", newx)
         self.setX(newx)
-        self.signals.windowMoved.emit()
+        # self.signals.windowMoved.emit()
 
     def setWindowWidth(self, newwidth):
         logging.debug("timeline.window: update window width to %f", newwidth)
@@ -38,7 +38,7 @@ class Window(QGraphicsRectItem):
         r = self.rect()
         r.setWidth(newwidth)
         self.setRect(r)
-        self.signals.windowMoved.emit()
+        # self.signals.windowMoved.emit()
 
     def setWindow(self, newx:float, newwidth:float):
         """
@@ -58,13 +58,14 @@ class Window(QGraphicsRectItem):
         r = self.rect()
         self.setRect(newx, r.y(), self._width, r.height())
         self.update()
-        self.signals.windowMoved.emit()
+        # self.signals.windowMoved.emit()
 
     def mouseMoveEvent(self, event):
         super().mouseMoveEvent(event)
 
     def mousePressEvent(self, event):
         self.setCursor(Qt.ClosedHandCursor)
+        # print(event.pos())
         super().mousePressEvent(event)
 
     def mouseReleaseEvent(self, event):
@@ -77,7 +78,7 @@ class Window(QGraphicsRectItem):
         if r.y() != self._y:
             self.setY(self._y)
         super().mouseReleaseEvent(event)
-        self.signals.windowMoved.emit()
+        self.signals.manualMove.emit()
 
     def hoverEnterEvent(self, event):
         super().hoverEnterEvent(event)
@@ -116,11 +117,12 @@ class DetectionTimeline(QWidget):
         font.setBold(True)
 
         self._window = Window(0, 0, 100, 60, window_pen, transparent_brush)
-        self._window.signals.windowMoved.connect(self.on_windowMoved)
+        self._window.signals.manualMove.connect(self.on_windowMoved)
 
         self._scene = QGraphicsScene(QRectF(0, 0, self._total_width, 85.))
         self._scene.setBackgroundBrush(self._bg_brush)
         self._scene.addItem(self._window)
+        self._scene.mousePressEvent = self.on_sceneMousePress
 
         self._view = QGraphicsView()
         # self._view.setRenderHints(QPainter.RenderHint.Antialiasing | QPainter.RenderHint.SmoothPixmapTransform)
@@ -151,12 +153,22 @@ class DetectionTimeline(QWidget):
         self._position_label.setFont(f)
 
         layout = QVBoxLayout()
+        layout.setSpacing(0)
+        layout.setContentsMargins(5, 2, 5, 2)
         layout.addWidget(self._view)
         layout.addWidget(self._position_label, Qt.AlignmentFlag.AlignRight)
         self.setLayout(layout)
         # self.setMaximumHeight(100)
         # self.setSizePolicy(QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Fixed)
 
+    def on_sceneMousePress(self, event):
+        scene_pos = event.scenePos()
+        relpos = scene_pos.x() / self._total_width
+        relpos = 0 if relpos < 0.0 else relpos
+        relpos = 2000/self._total_width if scene_pos.x() > self._total_width else relpos
+        self.signals.moveRequest.emit(relpos)
+        logging.debug("Timeline: Scene clicked at position: %.2f, %.2f --> rel x-pos %.3f", scene_pos.x(), scene_pos.y(), relpos)
+    
     def clear(self):
         for i in self._scene.items():
             if isinstance(i, (QGraphicsLineItem, QGraphicsEllipseItem)):
@@ -166,6 +178,7 @@ class DetectionTimeline(QWidget):
         logging.debug("Timeline: setData!")
         self._data = data
         self.update()
+        self.resizeEvent(None)
 
     def update(self):
         self.clear()
@@ -309,8 +322,7 @@ def main():
     datafile = PACKAGE_ROOT / "data/merged_small.pkl"
     with open(datafile, "rb") as f:
         df = pickle.load(f)
-    data = TrackingData()
-    data.setData(as_dict(df))
+    data = TrackingData(as_dict(df))
     data.setSelection(np.arange(0,100, 1))
     data.setUserLabeledStatus(True)
     start_x = 0.1
@@ -329,12 +341,14 @@ def main():
     backBtn.clicked.connect(lambda: back(0.2))
 
     btnLyt = QHBoxLayout()
+    btnLyt.setSpacing(1)
     btnLyt.addWidget(backBtn)
     btnLyt.addWidget(zeroBtn)
     btnLyt.addWidget(fwdBtn)
 
     view.setWindowPos(start_x)
     layout = QVBoxLayout()
+    layout.setSpacing(1)
     layout.addWidget(view)
     layout.addLayout(btnLyt)
     window.setLayout(layout)

fixtracks/widgets/detectionview.py

@@ -10,6 +10,7 @@ from fixtracks.utils.signals import DetectionSignals, DetectionViewSignals, Dete
 from fixtracks.utils.enums import DetectionData, Tracks
 from fixtracks.utils.trackingdata import TrackingData
 
+
 class Detection(QGraphicsEllipseItem):
     signals = DetectionSignals()
 
@@ -128,20 +129,23 @@ class DetectionView(QWidget):
                     del it
 
     def updateDetections(self, keypoint=-1):
+        logging.info("DetectionView.updateDetections!")
         self.clearDetections()
         if self._data is None:
             return
         frames = self._data.selectedData("frame")
         tracks = self._data.selectedData("track")
+        ids = self._data.selectedData("index")
         coordinates = self._data.coordinates(selection=True)
         centercoordinates = self._data.centerOfGravity(selection=True)
         userlabeled = self._data.selectedData("userlabeled")
-        indices = self._data.selectedData("index")
-        image_rect = self._pixmapitem.boundingRect() if self._pixmapitem is not None else QRectF(0,0,0,0)
+        scores = self._data.selectedData("confidence")
 
-        for i, idx in enumerate(indices):
-            t = tracks[i]
+        image_rect = self._pixmapitem.boundingRect() if self._pixmapitem is not None else QRectF(0,0,0,0)
+        num_detections = len(frames)
+        for i, (id, f, t, l, s) in enumerate(zip(ids, frames, tracks, userlabeled, scores)):
             c = Tracks.fromValue(t).toColor()
+            c.setAlpha(int(i * 255 / num_detections))
             if keypoint >= 0:
                 x = coordinates[i, keypoint, 0]
                 y = coordinates[i, keypoint, 1]
@@ -150,10 +154,12 @@ class DetectionView(QWidget):
                 y = centercoordinates[i, 1]
 
             item = Detection(image_rect.left() + x, image_rect.top() + y, 20, 20, brush=QBrush(c))
-            item.setData(DetectionData.TRACK_ID.value, tracks[i])
-            item.setData(DetectionData.ID.value, idx)
+            item.setData(DetectionData.TRACK_ID.value, t)
+            item.setData(DetectionData.ID.value, id)
             item.setData(DetectionData.COORDINATES.value, coordinates[i, :, :])
-            item.setData(DetectionData.FRAME.value, frames[i])
+            item.setData(DetectionData.FRAME.value, f)
+            item.setData(DetectionData.USERLABELED.value, l)
+            item.setData(DetectionData.SCORE.value, s)
             item = self._scene.addItem(item)
 
     def fit_image_to_view(self):
@@ -212,7 +218,7 @@ def main():
     view.setImage(img)
     view.addDetections(bg_coords, bg_tracks, bg_ids, background_brush)
     view.addDetections(focus_coords, focus_tracks, focus_ids, focus_brush)
-    view.addDetections(scnd_coords, scnd_tracks,  scnd_ids, second_brush)
+    view.addDetections(scnd_coords, scnd_tracks, scnd_ids, second_brush)
     window.setLayout(layout)
     window.show()
     app.exec()

fixtracks/widgets/selection_control.py

@@ -4,7 +4,7 @@ import numpy as np
 from PySide6.QtCore import Qt, Signal, QSize
 from PySide6.QtGui import QFont
 from PySide6.QtWidgets import QWidget, QLabel, QPushButton, QSizePolicy
-from PySide6.QtWidgets import QGridLayout, QVBoxLayout
+from PySide6.QtWidgets import QGridLayout, QVBoxLayout, QApplication
 
 from fixtracks.utils.styles import pushBtnStyle
 
@@ -15,6 +15,7 @@ class SelectionControls(QWidget):
     assignTwo = Signal()
     assignOther = Signal()
     accept = Signal()
+    accept_until = Signal()
     unaccept = Signal()
     delete = Signal()
     revertall = Signal()
@@ -51,7 +52,6 @@ class SelectionControls(QWidget):
         quarterstepBackBtn.setStyleSheet(pushBtnStyle("darkgray"))
         quarterstepBackBtn.clicked.connect(lambda: self.on_Back(quarterstep))
 
-
         fwdBtn = QPushButton(">>|")
         fwdBtn.setSizePolicy(QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Expanding)
         fwdBtn.setShortcut(Qt.Key.Key_Right)
@@ -102,7 +102,7 @@ class SelectionControls(QWidget):
         acceptBtn.setFont(font)
         acceptBtn.setSizePolicy(QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Expanding)
         acceptBtn.setStyleSheet(pushBtnStyle("darkgray"))
-        acceptBtn.setToolTip(f"Accept assignments of current selection as TRUE")
+        acceptBtn.setToolTip(f"Accept assignments of current selection as TRUE, Hold shift while clicking to accept all until here.")
         acceptBtn.clicked.connect(self.on_Accept)
 
         unacceptBtn = QPushButton("un-accept")
@@ -117,8 +117,9 @@ class SelectionControls(QWidget):
         deleteBtn.setSizePolicy(QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Expanding)
         deleteBtn.setStyleSheet(pushBtnStyle("red"))
         deleteBtn.setToolTip(f"DANGERZONE! Delete current selection of detections!")
-        deleteBtn.setEnabled(False)
+        deleteBtn.setShortcut("Ctrl+D")
         deleteBtn.clicked.connect(self.on_Delete)
+        deleteBtn.setEnabled(False)
 
         revertBtn = QPushButton("revert assignments")
         revertBtn.setFont(font)
@@ -171,7 +172,7 @@ class SelectionControls(QWidget):
         grid.setColumnStretch(0, 1)
         grid.setColumnStretch(7, 1)
         self.setLayout(grid)
-        self.setMaximumSize(QSize(500, 500))
+        # self.setMaximumSize(QSize(500, 500))
 
     def setWindow(self, start:int=0, end:int=0):
         self.startframe.setText(f"{start:.0f}")
@@ -210,7 +211,12 @@ class SelectionControls(QWidget):
 
     def on_Accept(self):
         logging.debug("SelectionControl: accept AssignmentBtn")
-        self.accept.emit()
+        modifiers = QApplication.keyboardModifiers()
+        if modifiers == Qt.KeyboardModifier.ShiftModifier:
+            logging.debug("Shift key was pressed during accept")
+            self.accept_until.emit()
+        else:
+            self.accept.emit()
 
     def on_Unaccept(self):
         logging.debug("SelectionControl: revoke user assignmentBtn")

fixtracks/widgets/skeleton.py

@@ -94,7 +94,8 @@ class SkeletonWidget(QWidget):
             i = s.data(DetectionData.ID.value)
             t = s.data(DetectionData.TRACK_ID.value)
             f = s.data(DetectionData.FRAME.value)
-            self._info_label.setText(f"Id {i}, track {t} on frame {f}, length {l:.1f} px")
+            sc = s.data(DetectionData.SCORE.value)
+            self._info_label.setText(f"Id {i}, track {t} on frame {f}, length {l:.1f} px, confidence {sc:.2f}")
         else:
             self._info_label.setText("")
 
@@ -129,7 +130,7 @@ class SkeletonWidget(QWidget):
         self._scene.setSceneRect(self._minx, self._miny, self._maxx - self._minx, self._maxy - self._miny)
         self._view.fitInView(self._scene.sceneRect(), Qt.AspectRatioMode.KeepAspectRatio)
 
-    def addSkeleton(self, coords, detection_id, frame, track, brush, update=True):
+    def addSkeleton(self, coords, detection_id, frame, track, score, brush, update=True):
         def check_extent(x, y, w, h):
             if x == 0 and y == 0:
                 return
@@ -157,12 +158,14 @@ class SkeletonWidget(QWidget):
         item.setData(DetectionData.ID.value, detection_id)
         item.setData(DetectionData.TRACK_ID.value, track)
         item.setData(DetectionData.FRAME.value, frame)
+        item.setData(DetectionData.SCORE.value, score)
         self._skeletons.append(item)
         if update:
             self.update()
 
     def addSkeletons(self, coordinates:np.ndarray, detection_ids:np.ndarray,
-                     frames:np.ndarray, tracks:np.ndarray, brush:QBrush):
+                     frames:np.ndarray, tracks:np.ndarray, scores:np.ndarray,
+                     brush:QBrush):
         num_detections = 0 if coordinates is None else coordinates.shape[0]
         logging.debug("SkeletonWidget: add %i Skeletons", num_detections)
         if num_detections < 1:
@@ -172,9 +175,10 @@ class SkeletonWidget(QWidget):
         detection_ids = detection_ids[sorting]
         frames = frames[sorting]
         tracks = tracks[sorting]
+        scores = scores[sorting]
         for i in range(num_detections):
             self.addSkeleton(coordinates[i,:,:], detection_ids[i], frames[i],
-                             tracks[i], brush=brush, update=False)
+                             tracks[i], scores[i], brush=brush, update=False)
         self.update()
 
     # def addSkeleton(self, coords, detection_id, brush):

fixtracks/widgets/tracks.py

@@ -1,11 +1,11 @@
 import logging
 import numpy as np
-import pandas as pd
 
 from PySide6.QtCore import Qt, QThreadPool, Signal
 from PySide6.QtGui import QImage, QBrush, QColor
 from PySide6.QtWidgets import QWidget, QVBoxLayout, QHBoxLayout, QLabel, QPushButton, QSizePolicy, QComboBox
-from PySide6.QtWidgets import QSpinBox, QSpacerItem, QProgressBar, QSplitter, QGridLayout, QFileDialog, QGridLayout
+from PySide6.QtWidgets import QSpinBox, QSpacerItem, QProgressBar, QSplitter, QFileDialog, QMessageBox
+
 
 from fixtracks.utils.reader import PickleLoader
 from fixtracks.utils.writer import PickleWriter
@@ -31,17 +31,12 @@ class FixTracks(QWidget):
         self._currentWindowPos = 0  # in frames
         self._currentWindowWidth = 0  # in frames
         self._maxframes = 0
+        self._manualmove = False
         self._data = None
 
         self._detectionView = DetectionView()
         self._detectionView.signals.itemsSelected.connect(self.on_detectionsSelected)
         self._skeleton = SkeletonWidget()
-        # self._skeleton.setMaximumSize(QSize(400, 400))
-        top_splitter = QSplitter(Qt.Orientation.Horizontal)
-        top_splitter.addWidget(self._detectionView)
-        top_splitter.addWidget(self._skeleton)
-        top_splitter.setStretchFactor(0, 2)
-        top_splitter.setStretchFactor(1, 1)
 
         self._progress_bar = QProgressBar(self)
         self._progress_bar.setMaximumHeight(20)
@@ -50,6 +45,7 @@ class FixTracks(QWidget):
 
         self._timeline = DetectionTimeline()
         self._timeline.signals.windowMoved.connect(self.on_windowChanged)
+        self._timeline.signals.moveRequest.connect(self.on_moveRequest)
 
         self._windowspinner = QSpinBox()
         self._windowspinner.setRange(10, 10000)
@@ -60,15 +56,31 @@ class FixTracks(QWidget):
         self._keypointcombo = QComboBox()
         self._keypointcombo.currentIndexChanged.connect(self.on_keypointSelected)
 
-        combo_layout = QGridLayout()
-        combo_layout.addWidget(QLabel("Window:"), 0, 0)
-        combo_layout.addWidget(self._windowspinner, 0, 1)
-        combo_layout.addWidget(QLabel("Keypoint:"), 1, 0)
-        combo_layout.addWidget(self._keypointcombo, 1, 1)
+        self._goto_spinner = QSpinBox()
+        self._goto_spinner.setSingleStep(1)
 
-        timelinebox = QHBoxLayout()
-        timelinebox.addWidget(self._timeline)
+        self._gotobtn = QPushButton("go!")
+        self._gotobtn.setToolTip("Jump to a given frame")
+        self._gotobtn.clicked.connect(self.on_goto)
+
+        combo_layout = QHBoxLayout()
+        combo_layout.addWidget(QLabel("Window width:"))
+        combo_layout.addWidget(self._windowspinner)
+        combo_layout.addWidget(QLabel("frames"))
+        combo_layout.addItem(QSpacerItem(10, 10, QSizePolicy.Policy.Fixed, QSizePolicy.Policy.Fixed))
+        combo_layout.addWidget(QLabel("Keypoint:"))
+        combo_layout.addWidget(self._keypointcombo)
+        combo_layout.addItem(QSpacerItem(10, 10, QSizePolicy.Policy.Fixed, QSizePolicy.Policy.Fixed))
+        combo_layout.addWidget(QLabel("Jump to frame:"))
+        combo_layout.addWidget(self._goto_spinner)
+        combo_layout.addWidget(self._gotobtn)
+        combo_layout.addItem(QSpacerItem(100, 10, QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Fixed))
+        combo_layout.setSpacing(1)
+
+        timelinebox = QVBoxLayout()
+        timelinebox.setSpacing(2)
         timelinebox.addLayout(combo_layout)
+        timelinebox.addWidget(self._timeline)
 
         self._controls_widget = SelectionControls()
         self._controls_widget.assignOne.connect(self.on_assignOne)
@@ -77,6 +89,7 @@ class FixTracks(QWidget):
         self._controls_widget.fwd.connect(self.on_forward)
         self._controls_widget.back.connect(self.on_backward)
         self._controls_widget.accept.connect(self.on_setUserFlag)
+        self._controls_widget.accept_until.connect(self.on_setUserFlagsUntil)
         self._controls_widget.unaccept.connect(self.on_unsetUserFlag)
         self._controls_widget.delete.connect(self.on_deleteDetection)
         self._controls_widget.revertall.connect(self.on_revertUserFlags)
@@ -102,6 +115,7 @@ class FixTracks(QWidget):
         data_selection_box.addWidget(QLabel("Select data file"))
         data_selection_box.addWidget(self._data_combo)
         data_selection_box.addItem(QSpacerItem(100, 10, QSizePolicy.Policy.Expanding, QSizePolicy.Policy.Fixed))
+        data_selection_box.setSpacing(0)
 
         btnBox = QHBoxLayout()
         btnBox.setAlignment(Qt.AlignmentFlag.AlignLeft)
@@ -117,9 +131,14 @@ class FixTracks(QWidget):
         cntrlBox = QHBoxLayout()
         cntrlBox.addWidget(self._classifier)
         cntrlBox.addWidget(self._controls_widget, alignment=Qt.AlignmentFlag.AlignCenter)
-        cntrlBox.addItem(QSpacerItem(300, 100, QSizePolicy.Policy.Fixed, QSizePolicy.Policy.Expanding))
+        cntrlBox.addWidget(self._skeleton)
+        cntrlBox.addItem(QSpacerItem(50, 100, QSizePolicy.Policy.Fixed, QSizePolicy.Policy.Expanding))
+        cntrlBox.setSpacing(0)
+        cntrlBox.setContentsMargins(0,0,0,0)
 
         vbox = QVBoxLayout()
+        vbox.setSpacing(0)
+        vbox.setContentsMargins(0,0,0,0)
         vbox.addLayout(timelinebox)
         vbox.addLayout(cntrlBox)
         vbox.addLayout(btnBox)
@@ -127,13 +146,16 @@ class FixTracks(QWidget):
         container.setLayout(vbox)
 
         splitter = QSplitter(Qt.Orientation.Vertical)
-        splitter.addWidget(top_splitter)
+        splitter.addWidget(self._detectionView)
         splitter.addWidget(container)
         splitter.setStretchFactor(0, 3)
         splitter.setStretchFactor(1, 1)
+
         layout = QVBoxLayout()
         layout.addLayout(data_selection_box)
         layout.addWidget(splitter)
+        layout.setSpacing(0)
+        layout.setContentsMargins(5,2,2,5)
         self.setLayout(layout)
 
     def on_autoClassify(self, tracks):
@@ -159,14 +181,19 @@ class FixTracks(QWidget):
         self._detectionView.setImage(img)
 
     def update(self):
+        kp = self._keypointcombo.currentText().lower()
+        if len(kp) == 0:
+            return
+        kpi = -1 if "center" in kp else int(kp)
+
         start_frame = self._currentWindowPos
         stop_frame = start_frame + self._currentWindowWidth
+
+        self._timeline.setWindow(start_frame / self._maxframes,
+                                 self._currentWindowWidth/self._maxframes)
         logging.debug("Tracks:update: Updating View for detection range %i, %i frames", start_frame, stop_frame)
         self._data.setSelectionRange("frame", start_frame, stop_frame)
-
         self._controls_widget.setWindow(start_frame, stop_frame)
-        kp = self._keypointcombo.currentText().lower()
-        kpi = -1 if "center" in kp else int(kp)
         self._detectionView.updateDetections(kpi)
 
     @property
@@ -208,11 +235,12 @@ class FixTracks(QWidget):
             self._saveBtn.setEnabled(True)
             self._currentWindowPos = 0
             self._currentWindowWidth = self._windowspinner.value()
-            self._maxframes = self._data.max("frame")
+            self._maxframes = np.max(self._data["frame"])
+            self._goto_spinner.setMaximum(self._maxframes)
             self.populateKeypointCombo(self._data.numKeypoints())
             self._timeline.setData(self._data)
-            self._timeline.setWindow(self._currentWindowPos / self._maxframes,
-                                     self._currentWindowWidth / self._maxframes)
+            # self._timeline.setWindow(self._currentWindowPos / self._maxframes,
+            #                          self._currentWindowWidth / self._maxframes)
             self._detectionView.setData(self._data)
             self._classifier.setData(self._data)
             self.update()
@@ -268,6 +296,12 @@ class FixTracks(QWidget):
         self._timeline.update()
         self.update()
 
+    def on_setUserFlagsUntil(self):
+        self._data.setSelectionRange("frame", 0, self._currentWindowPos + self._currentWindowWidth)
+        self._data.setUserLabeledStatus(True)
+        self._timeline.update()
+        self.update()
+
     def on_unsetUserFlag(self):
         logging.debug("Tracks:unsetUserFlag")
         self._data.setUserLabeledStatus(False)
@@ -276,20 +310,43 @@ class FixTracks(QWidget):
 
     def on_revertUserFlags(self):
         logging.debug("Tracks:revert ALL UserFlags and track assignments")
-        self._data.revertAssignmentStatus()
-        self._data.revertTrackAssignments()
+        msg_box = QMessageBox()
+        msg_box.setIcon(QMessageBox.Icon.Warning)
+        msg_box.setText(f"Are you sure you want to revert ALL track assignments?")
+        msg_box.setStandardButtons(QMessageBox.StandardButton.Yes | QMessageBox.StandardButton.No)
+        msg_box.setDefaultButton(QMessageBox.StandardButton.No)
+        ret = msg_box.exec()
+
+        if ret == QMessageBox.StandardButton.Yes:
+            self._data.revertUserLabeledStatus()
+            self._data.revertTrackAssignments()
         self._timeline.update()
         self.update()
 
     def on_deleteDetection(self):
-        logging.warning("Tracks:delete detections is currently not supported!")
-        # self._data.deleteDetections()
+        logging.info("Tracks:deleting detections!")
+        msg_box = QMessageBox()
+        msg_box.setIcon(QMessageBox.Icon.Warning)
+        msg_box.setText(f"Are you sure you want to delete the selected ({len(self._data.selectionIndices)})detections?")
+        msg_box.setStandardButtons(QMessageBox.StandardButton.Yes | QMessageBox.StandardButton.No)
+        msg_box.setDefaultButton(QMessageBox.StandardButton.No)
+        ret = msg_box.exec()
+
+        if ret == QMessageBox.StandardButton.Yes:
+            self._data.deleteDetections()
         self._timeline.update()
         self.update()
 
     def on_windowChanged(self):
         logging.debug("Tracks:Timeline reports window change ")
-        self._currentWindowPos = np.round(self._timeline.rangeStart * self._maxframes)
+        if not self._manualmove:
+            self._currentWindowPos = np.round(self._timeline.rangeStart * self._maxframes)
+            self.update()
+        self._manualmove = False
+
+    def on_moveRequest(self, pos):
+        new_pos = int(np.round(pos * self._maxframes))
+        self._currentWindowPos = new_pos
         self.update()
 
     def on_windowSizeChanged(self, value):
@@ -302,14 +359,29 @@ class FixTracks(QWidget):
         """
         self._currentWindowWidth = value
         logging.debug("Tracks:OnWindowSizeChanged %i franes", value)
-        self._timeline.setWindowWidth(self._currentWindowWidth / self._maxframes)
-        self._controls_widget.setSelectedTracks(None)
+        # if self._maxframes == 0:
+        #     self._timeline.setWindowWidth(self._currentWindowWidth / 2000)
+        # else:
+        #     self._timeline.setWindowWidth(self._currentWindowWidth / self._maxframes)
+        # self._controls_widget.setSelectedTracks(None)
+        self.update()
+
+    def on_goto(self):
+        target = self._goto_spinner.value()
+        if target > self._maxframes - self._currentWindowWidth:
+           target =  self._maxframes - self._currentWindowWidth
+        logging.info("Jump to frame %i", target)
+        self._currentWindowPos = target
+        self._timeline.setWindow(self._currentWindowPos / self._maxframes, 
+                                 self._currentWindowWidth / self._maxframes)
+        self.update()
 
     def on_detectionsSelected(self, detections):
-        logging.debug("Tracks: Detections selected")
+        logging.debug("Tracks: %i Detections selected", len(detections))
         tracks = np.zeros(len(detections), dtype=int)
         ids = np.zeros_like(tracks)
         frames = np.zeros_like(tracks)
+        scores = np.zeros(tracks.shape, dtype=float)
         coordinates = None
         if len(detections) > 0:
             c = detections[0].data(DetectionData.COORDINATES.value)
@@ -320,16 +392,20 @@ class FixTracks(QWidget):
             ids[i] = d.data(DetectionData.ID.value)
             frames[i] = d.data(DetectionData.FRAME.value)
             coordinates[i, :, :] = d.data(DetectionData.COORDINATES.value)
+            scores[i] = d.data(DetectionData.SCORE.value)
         self._data.setSelection(ids)
         self._controls_widget.setSelectedTracks(tracks)
         self._skeleton.clear()
-        self._skeleton.addSkeletons(coordinates, ids, frames, tracks, QBrush(QColor(10, 255, 65, 255)))
-        self.update()
+        self._skeleton.addSkeletons(coordinates, ids, frames, tracks, scores, QBrush(QColor(10, 255, 65, 255)))
 
     def moveWindow(self, stepsize):
-        step = np.round(stepsize * (self._currentWindowWidth))
-        new_start_frame = self._currentWindowPos + step
-        self._timeline.setWindowPos(new_start_frame / self._maxframes)
+        logging.info("Tracks.moveWindow: move window with stepsize %.2f", stepsize)
+        self._manualmove = True
+        new_start_frame = self._currentWindowPos + np.round(stepsize * self._currentWindowWidth)
+        if new_start_frame < 0:
+            new_start_frame = 0
+        elif new_start_frame + self._currentWindowWidth > self._maxframes:
+            new_start_frame = self._maxframes - self._currentWindowWidth
         self._currentWindowPos = new_start_frame
         self._controls_widget.setSelectedTracks(None)
         self.update()