Mercity/Figure-Skating-Classification-Data
AI Skating Coach - Figure Skating Element Recognition Dataset Clean 64-class version with multi-jump combinations preserved Overview Figure skating skeleton pose sequences for action/element classification. Raw keypoint data extracted from competition videos and professional motion capture, presented in clean unmodified form. Total samples: 5,405 Training: 4,324 sequences Test: 1,081 sequences Classes: 64 figure skating elements Format: Clean unaugmented data… See the full description on the dataset page: https://huggingface.co/datasets/Mercity/Figure-Skating-Classification-Data.
052
1{2 "dataset_name": "AI Skating Coach - Clean 64-Class",3 "version": "1.0",4 "description": "Figure skating element recognition with 64 classes including multi-jump combinations",5 "total_samples": 5168,6 "train_samples": 4161,7 "test_samples": 1007,8 "split_ratio": "80/20",9 "num_classes": 64,10 "sequence_length": 150,11 "num_keypoints": 17,12 "coordinates_per_keypoint": 3,13 "duration_seconds": 5.0,14 "fps": 30,15 "keypoint_format": "COCO 17-point",16 "data_characteristics": {17 "no_augmentation": true,18 "no_class_weights": true,19 "no_synthetic_samples": true,20 "combinations_preserved": true,21 "format": "Pickle (NumPy-compatible, list of arrays)"22 },23 "data_sources": [24 "MMFS Dataset: 2D pose from competition videos (4,915 sequences)",25 "JSON Motion Capture: Professional 3D mocap (253 sequences)"26 ],27 "preprocessing_steps": [28 "Format unification (142-marker mocap \u2192 17-keypoint COCO)",29 "Temporal uniform sampling to 150 frames",30 "Coordinate normalization to [-1, 1]",31 "Velocity feature computation",32 "80/20 stratified train/test split"33 ],34 "keypoint_definitions": {35 "0-4": "Head (nose, left_eye, right_eye, left_ear, right_ear)",36 "5-10": "Upper body (left_shoulder, right_shoulder, left_elbow, right_elbow, left_wrist, right_wrist)",37 "11-16": "Lower body (left_hip, right_hip, left_knee, right_knee, left_ankle, right_ankle)"38 }39}