Ready-made dataset

High-Quality Dancing Pose Dataset (20 videos)

Train your AI faster with clean, ready-to-use motion datasets.
No preprocessing. No noise. Just results.

This is a clean, processed Dancing Motion Pose Dataset containing 11,886 HQ frames (including augmentations), extracted from 20 different dancing videos.

One-time: $250Action: dancingFormat: JSONL

Key Features

  • Total Frames: 11,886 HQ frames (includes augmentations)
  • Original Frames: 3,962 strictly filtered HQ frames (pre-augmentation)
  • Processing: Gaussian temporal smoothing (window=5), hip-center body normalization, horizontal flip + keypoint noise augmentation
  • Pose Format: Full 33 keypoints (x, y, z + visibility)
  • Additional Data: Sequence features (velocity/acceleration proxies, step frequency, motion consistency), per-frame quality scores, and full quality reports

Quality Metrics (Accepted Frames)

  • Mean Quality Score: 0.8730
  • Mean Landmark Visibility: 0.8360
  • Acceptance Rate: 79.54% (pre-aug / sampled)

Structured data layout

  • data.jsonl — merged export (accepted frames)
  • per_video/*.jsonl — per-video splits + rejected per-video files
  • features.json — sequence-level motion features (velocity/acceleration proxies)
  • low_quality_frames.jsonl — rejected frames for hard-example mining
  • export_quality_report.json — detailed quality diagnostics
  • manifest.json & global_stats.json — metadata and aggregates
  • SCHEMA.md + examples/ — schema and sample rows

Free samples

Validate the format on GitHub: QualityVision-Motion-Dataset-Samples.

Purchase & licensing

  • Checkout: Gumroad (one-time $250; see product page)
  • Deliverable: JSONL + metadata exports — not raw source videos
  • License: Proprietary / commercial terms as listed on Gumroad
  • Contact: info@qvision.space
Buy on Gumroad $250 Contact for custom datasetsinfo@qvision.spaceBack to pricing