Modeling and analysis of jumps among elite figure skaters

<div> <div><strong>Modeling and Analysis of Jumps Among Elite Figure Skaters</strong></div> <br> <div>This repository contains:</div> <br> <ol> <li><strong>3A.mp4</strong>: An example video clip of a jump.</li> <li><strong>3A_skeleton_overlay.mp4</strong>: A skeleton overlay of the jump used to extract key points.</li> <li><strong>data.csv</strong>: A CSV file containing data on 20 jumps (10 male, 10 female) performed during the 2022 Winter Olympics. The data includes the following columns:</li> </ol> <ul> <li> <ul> <li>gender</li> <li>jump_type: The jump type as recorded by the judges (<a href="https://results.isu.org/results/season2122/owg2022/">link to results</a>)</li> <li>jump_revolutions: Number of revolutions in the jump.</li> <li>jump_rotations: Number of degrees of rotation in the jump.</li> <li>jump_classification: ISU abbreviation of the jump (T: Toe Loop, F: Flip, Lz: Lutz, S: Salchow, L: Loop, A: Axel).</li> <li>frames_in_air: Number of frames the athlete was in the air, based on a 30 FPS broadcast.</li> <li>time_in_air: Seconds the athlete was in the air.</li> <li>takeoff_leg: Last leg on the ice at the start of the jump.</li> <li>landing_leg: First leg to touch the ice upon landing.</li> <li>jump_rotations_in_air: Manually identified degrees of rotation performed in the air (from takeoff to landing).</li> </ul> </li> </ul> <br> <div>This data has been anonymized using k-anonymity, with k = 3.</div> </div>