fixed args

Signed-off-by: Mpho Mphego <mpho112@gmail.com>

Author: mmphego

main.py

@@ -97,14 +97,20 @@ def arg_parser():
         "--mouse-precision",
         type=str,
         default="low",
-        help="The precision for mouse movement (how much the mouse moves).",
+        const="low",
+        nargs="?",
+        choices=["high", "low", "medium"],
+        help="The precision for mouse movement (how much the mouse moves). [Default: low]",
     )
     parser.add_argument(
         "-ms",
         "--mouse-speed",
         type=str,
         default="fast",
-        help="The speed (how fast it moves) by changing",
+        const="fast",
+        nargs="?",
+        choices=["fast", "slow", "medium"],
+        help="The speed (how fast it moves) by changing [Default: fast]",
     )
     parser.add_argument(
         "--enable-mouse", action="store_true", help="Enable Mouse Movement",
@@ -113,7 +119,9 @@ def arg_parser():
         "--debug", action="store_true", help="Show output on screen [debugging].",
     )
     parser.add_argument(
-        "--show-bbox", action="store_true", help="Show bounding box and stats on screen [debugging].",
+        "--show-bbox",
+        action="store_true",
+        help="Show bounding box and stats on screen [debugging].",
     )
 
     return parser.parse_args()
@@ -148,7 +156,9 @@ def main(args):
 
     for frame in video_feed.next_frame():
 
-        predict_end_time, face_bboxes = face_detection.predict(frame, show_bbox=args.show_bbox)
+        predict_end_time, face_bboxes = face_detection.predict(
+            frame, show_bbox=args.show_bbox
+        )
 
         if face_bboxes:
             for face_bbox in face_bboxes:
@@ -163,7 +173,7 @@ def main(args):
                 (x, y, w, h) = face_bbox
                 face = frame[y:h, x:w]
                 (face_height, face_width) = face.shape[:2]
-                #  video_feed.show(frame[y:h, x:w], "face")
+                # video_feed.show(frame[y:h, x:w], "face")
 
                 # ensure the face width and height are sufficiently large
                 if face_height < 20 or face_width < 20:
@@ -188,9 +198,9 @@ def main(args):
                     head_pose_estimation.show_text(frame, head_pose_angles)
                     gaze_estimation.show_text(frame, gaze_vector)
 
-                print(f"gaze_vector: {gaze_vector}")
+                print(f"gaze_vector: x: {gaze_vector['x']:.2f}, y: {gaze_vector['y']:.2f}")
                 if args.enable_mouse:
-                    mouse_controller.move(gaze_vector['x'], gaze_vector['y'])
+                    mouse_controller.move(gaze_vector["x"], gaze_vector["y"])
 
         if args.debug:
             video_feed.show(video_feed.resize(frame))

src/model.py

@@ -254,22 +254,23 @@ def preprocess_output(self, inference_results, image, show_bbox=False):
         flattened_predictions = np.vstack(inference_results).ravel()
         eyes_coords = flattened_predictions[:4]
         h, w = image.shape[:2]
+        eye_size = 10
 
         left_eye_x_coord = int(eyes_coords[0] * w)
-        left_eye_xmin = left_eye_x_coord - 10
-        left_eye_xmax = left_eye_x_coord + 10
+        left_eye_xmin = left_eye_x_coord - eye_size
+        left_eye_xmax = left_eye_x_coord + eye_size
 
         left_eye_y_coord = int(eyes_coords[1] * h)
-        left_eye_ymin = left_eye_y_coord - 10
-        left_eye_ymax = left_eye_y_coord + 10
+        left_eye_ymin = left_eye_y_coord - eye_size
+        left_eye_ymax = left_eye_y_coord + eye_size
 
         right_eye_x_coord = int(eyes_coords[2] * w)
-        right_eye_xmin = right_eye_x_coord - 10
-        right_eye_xmax = right_eye_x_coord + 10
+        right_eye_xmin = right_eye_x_coord - eye_size
+        right_eye_xmax = right_eye_x_coord + eye_size
 
         right_eye_y_coord = int(eyes_coords[3] * h)
-        right_eye_ymin = right_eye_y_coord - 10
-        right_eye_ymax = right_eye_y_coord + 10
+        right_eye_ymin = right_eye_y_coord - eye_size
+        right_eye_ymax = right_eye_y_coord + eye_size
 
         eyes_coords = {
             "left_eye_point": (left_eye_x_coord, left_eye_y_coord),
@@ -286,7 +287,7 @@ def preprocess_output(self, inference_results, image, show_bbox=False):
         return eyes_coords, image
 
     @staticmethod
-    def draw_output(image, eyes_coords, radius=10, color=(0, 0, 255), thickness=2):
+    def draw_output(image, eyes_coords, radius=20, color=(0, 0, 255), thickness=2):
         """Draw a circle around ROI"""
         for eye, coords in eyes_coords.items():
             if "point" in eye:
@@ -445,12 +446,11 @@ def preprocess_output(self, inference_results, image, show_bbox, **kwargs):
     def draw_output(coords, image, **kwargs):
         left_eye_point = kwargs["eyes_coords"]["left_eye_point"]
         right_eye_point = kwargs["eyes_coords"]["right_eye_point"]
-        print(left_eye_point)
         cv2.arrowedLine(
             image,
             (
                 left_eye_point[0] + int(coords["x"] * 500),
-                left_eye_point[1] + int(-coords["y"] * 500),
+                left_eye_point[1] - int(coords["y"] * 500),
             ),
             (left_eye_point[0], left_eye_point[1]),
             color=(0, 0, 255),
@@ -461,7 +461,7 @@ def draw_output(coords, image, **kwargs):
             image,
             (
                 right_eye_point[0] + int(coords["x"] * 500),
-                right_eye_point[1] + int(-coords["y"] * 500),
+                right_eye_point[1] - int(coords["y"] * 500),
             ),
             (right_eye_point[0], right_eye_point[1]),
             color=(0, 0, 255),

src/mouse_controller.py

@@ -25,14 +25,14 @@ def move(self, x, y):
         try:
             start_pos = pyautogui.position()
             pyautogui.moveRel(
-                x * self.precision, -1 * y * self.precision, duration=self.speed
+                -x * self.precision, y * self.precision, duration=self.speed
             )
             end_pos = pyautogui.position()
             logger.info(f"Mouse -> start_pos: {start_pos}, end_pos: {end_pos}")
         except pyautogui.FailSafeException:
             logger.exception(f"Position: {x}, {y} are out of the screen")
             pyautogui.moveRel(
-                -x * self.precision, 1 * y * self.precision, duration=self.speed
+                x * self.precision, -1 * y * self.precision, duration=self.speed
             )
     def left_click(self):
         pass