Implement pytorch dataset for own collected data

src/keypoint_extractor.py

@@ -1,12 +1,14 @@
-import mediapipe as mp
-import cv2
-import time
-from typing import Dict, List, Tuple
-import numpy as np
 import logging
 import os
+import time
+from typing import Dict, List, Tuple
+
+import cv2
+import mediapipe as mp
+import numpy as np
 import pandas as pd
 
+
 class KeypointExtractor:
     def __init__(self, video_folder: str, cache_folder: str = "cache"):
         self.mp_drawing = mp.solutions.drawing_utils
@@ -52,7 +54,18 @@ class KeypointExtractor:
 
         keypoints_df = pd.DataFrame(columns=self.columns)
 
+        # extract frames from video so we extract 5 frames per second
+        frame_rate = int(cap.get(cv2.CAP_PROP_FPS))
+        frame_skip = frame_rate // 5
+
         while cap.isOpened():
+
+            # skip frames
+            for _ in range(frame_skip):
+                success, image = cap.read()
+                if not success:
+                    break
+
             success, image = cap.read()
             if not success:
                 break