I would like to train a Model to predict the Actions in a Video based on Mediapipe Keypoints. Framenumber,Keypoint_Pose_0_X,Keypoint_Pose_0_Y,Keypoint_Pose_0_Visibility,Keypoint_Pose_1_X,Keypoint_Pose_1_Y,Keypoint_Pose_1_Visibility,Keypoint_Pose_2_X,Keypoint_Pose_2_Y,Keypoint_Pose_2_Visibility,Keypoint_Pose_3_X,Keypoint_Pose_3_Y,Keypoint_Pose_3_Visibility,Keypoint_Pose_4_X,Keypoint_Pose_4_Y,Keypoint_Pose_4_Visibility,Keypoint_Pose_5_X,Keypoint_Pose_5_Y,Keypoint_Pose_5_Visibility,Keypoint_Pose_6_X,Keypoint_Pose_6_Y,Keypoint_Pose_6_Visibility,Keypoint_Pose_7_X,Keypoint_Pose_7_Y,Keypoint_Pose_7_Visibility,Keypoint_Pose_8_X,Keypoint_Pose_8_Y,Keypoint_Pose_8_Visibility,Keypoint_Pose_9_X,Keypoint_Pose_9_Y,Keypoint_Pose_9_Visibility,Keypoint_Pose_10_X,Keypoint_Pose_10_Y,Keypoint_Pose_10_Visibility,Keypoint_Pose_11_X,Keypoint_Pose_11_Y,Keypoint_Pose_11_Visibility,Keypoint_Pose_12_X,Keypoint_Pose_12_Y,Keypoint_Pose_12_Visibility,Keypoint_Pose_13_X,Keypoint_Pose_13_Y,Keypoint_Pose_13_Visibility,Keypoint_Pose_14_X,Keypoint_Pose_14_Y,Keypoint_Pose_14_Visibility,Keypoint_Pose_15_X,Keypoint_Pose_15_Y,Keypoint_Pose_15_Visibility,Keypoint_Pose_16_X,Keypoint_Pose_16_Y,Keypoint_Pose_16_Visibility,Keypoint_Pose_17_X,Keypoint_Pose_17_Y,Keypoint_Pose_17_Visibility,Keypoint_Pose_18_X,Keypoint_Pose_18_Y,Keypoint_Pose_18_Visibility,Keypoint_Pose_19_X,Keypoint_Pose_19_Y,Keypoint_Pose_19_Visibility,Keypoint_Pose_20_X,Keypoint_Pose_20_Y,Keypoint_Pose_20_Visibility,Keypoint_Pose_21_X,Keypoint_Pose_21_Y,Keypoint_Pose_21_Visibility,Keypoint_Pose_22_X,Keypoint_Pose_22_Y,Keypoint_Pose_22_Visibility,Keypoint_Pose_23_X,Keypoint_Pose_23_Y,Keypoint_Pose_23_Visibility,Keypoint_Pose_24_X,Keypoint_Pose_24_Y,Keypoint_Pose_24_Visibility,Keypoint_Pose_25_X,Keypoint_Pose_25_Y,Keypoint_Pose_25_Visibility,Keypoint_Pose_26_X,Keypoint_Pose_26_Y,Keypoint_Pose_26_Visibility,Keypoint_Pose_27_X,Keypoint_Pose_27_Y,Keypoint_Pose_27_Visibility,Keypoint_Pose_28_X,Keypoint_Pose_28_Y,Keypoint_Pose_28_Visibility,Keypoint_Pose_29_X,Keypoint_Pose_29_Y,Keypoint_Pose_29_Visibility,Keypoint_Pose_30_X,Keypoint_Pose_30_Y,Keypoint_Pose_30_Visibility,Keypoint_Pose_31_X,Keypoint_Pose_31_Y,Keypoint_Pose_31_Visibility,Keypoint_Pose_32_X,Keypoint_Pose_32_Y,Keypoint_Pose_32_Visibility,Action 0,481,118,0.9999821186065674,489,106,0.9999444484710692,494,107,0.9998898506164552,498,107,0.9998225569725036,472,105,0.9999796152114868,465,104,0.9999828338623048,460,103,0.9999879598617554,502,114,0.9993197917938232,452,109,0.999990463256836,488,133,0.9998002648353576,469,132,0.9999774694442748,493,182,0.995231568813324,431,184,0.999991536140442,488,273,0.0262717884033918,401,281,0.9973281621932985,513,338,0.0641939714550972,459,362,0.984261393547058,520,357,0.0794331803917884,473,383,0.9602025747299194,525,355,0.0814166516065597,482,373,0.9599066376686096,520,349,0.0855123922228813,480,366,0.9528304934501648,497,370,0.998590648174286,476,375,0.9997231364250184,491,466,0.3849213421344757,487,468,0.9575459361076356,498,563,0.6474338173866272,493,563,0.8860534429550171,492,582,0.7143721580505371,485,582,0.7824604511260986,544,582,0.7379774451255798,537,596,0.8613807559013367,Action_1 This is the Head of my merged Dataset from different Videos i created with: import cv2 import mediapipe as mp import csv import time import random import logging import os import pandas as pd logging.basicConfig(level=logging.DEBUG, … Read more