datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
FineGym-skeleton
FineGym-skeleton Dataset
License: CC BY 4.0
Overview
FineGym-skeleton is a human-skeleton action-recognition benchmark derived from FineGym. It combines temporally precise, fine-grained gymnastics annotations with 2D human-pose sequences extracted from the corresponding video frames. The current V2 release contains Gym99-skeleton-V2 and Gym288-skeleton-V2. The source RGB subaction clips are also available in the FineGym-RGB-subactions directory on Hugging Face.… See the full description on the dataset page: https://huggingface.co/datasets/Lozumi/FineGym-skeleton.eu-hydro-master-skeleton
EU-Hydro Master Skeleton
Per-basin GeoParquet shards derived from the Copernicus EU-Hydro v1.3 GeoPackages. Four layers are published — river centerlines, river-surface polygons, inland-water polygons (lakes + wide waters), and river-basin polygons — all reprojected to a common CRS and stripped of admin-only columns for easier querying.
Contents
eu_hydro_master_skeleton_geoparquet/
├── river_lines/ # River_Net_l MultiLineString ~1.3 M features
├──… See the full description on the dataset page: https://huggingface.co/datasets/InfoVis-Project-Group-19/eu-hydro-master-skeleton.ANUBIS-Skeleton
ANUBIS Dataset
ANUBIS Dataset
Large-Scale Skeleton-Based Action Recognition
A comprehensive multi-view skeleton action dataset for challenging real-world scenarios
📄 Paper •
🌐 Project Website •
📊 Download Dataset •
💻 Benchmark Code
📝 Overview
ANUBIS is a large-scale skeleton-based action recognition dataset designed to address critical gaps in existing benchmarks. The dataset features 102 action categories collected from 80 participants… See the full description on the dataset page: https://huggingface.co/datasets/Khat865/ANUBIS-Skeleton.gaikotsu_no_yurei_skeleton_spectresarc-o-meter-exercise-skeleton
Sarc-O-Meter Exercise Skeleton Dataset
A 33-keypoint MediaPipe skeleton dataset for exercise movement analysis.
Exercises
Calf Raise
Sit to Stand
Step Up
Speed Classes
Slow
Normal
Fast
Dataset Format
The skeleton data is stored as .npy files. Each sequence contains MediaPipe Pose landmarks with the shape:
Frames × 33 × 4
The four values represent:
x
y
z
visibility
The dataset also contains metadata.parquet, which provides… See the full description on the dataset page: https://huggingface.co/datasets/Surya2212/sarc-o-meter-exercise-skeleton.fin7-skeletons
FIN7 skeletons — Ready ≠ measured
Live board: GET https://councilof.ai/api/gspc — 22 axis · 15 measured · 7 UNMEASURED.
These JSON shells are not scores. C-2026-0826-05 forbids restoring MEASURED-INDEX-v0.1. Humanoid = NOT_BUILT. We are not a Transparency Service.
Do not stamp MEASURED from a skeleton.
skeleton
Skeleton CLI Training Data
1,843 agent session traces exported via skeleton-cli.
Source
Rows
Description
opencode
724
OpenCode CLI agent sessions
hermes
258
Hermes agent sessions
rtk
846
RTK command execution logs
openclaw
14
OpenClaw agent sessions
kiro
1
Kiro agent session
SkeletonsData_Skeleton_Action_Recognitionuz-lexicon-skeleton
Uzbek frequency lexicon with skeleton index
English · Oʻzbekcha
Derived from tahrirchi/uz-books-v2 (MIT)
and tahrirchi/uz-crawl (Apache-2.0).
1.75 billion word tokens counted: 1.43 B from books, 296 M from news, 25.5 M from Telegram channels.
Licence: CC BY 4.0 — credit this dataset and both sources.
4,892,125 Uzbek word forms with their frequencies, 23,160,605 word pairs, and a
skeleton index that groups the words a keyboard makes indistinguishable. Every word is
written in… See the full description on the dataset page: https://huggingface.co/datasets/Maqsudjonpolatov/uz-lexicon-skeleton.spider-skeleton-context-instruct
Dataset Card for Spider Skeleton Context Instruct
Dataset Summary
Spider is a large-scale complex and cross-domain semantic parsing and text-to-SQL dataset annotated by 11 Yale students
The goal of the Spider challenge is to develop natural language interfaces to cross-domain databases.
This dataset was created to finetune LLMs in a ### Instruction: and ### Response: format with database context.
Yale Lily Spider Leaderboards
The leaderboard can be seen at… See the full description on the dataset page: https://huggingface.co/datasets/richardr1126/spider-skeleton-context-instruct.irds-skeleton-interpretation
IRDS Skeleton Interpretation Visualizations
Per-joint interpretation visualizations (animated 3D skeleton GIFs) and
region-concentration tables for deep models trained on the IntelliRehabDS
(IRDS) dataset — binary patient-vs-control classification and pose
forecasting from Kinect v2 skeletons (25 joints).
Contents
val_interp_gifs.zip (≈ 942 MB, 1684 GIFs)
Animated 3D-pose GIFs where each joint is colored by its interpretation
importance at each… See the full description on the dataset page: https://huggingface.co/datasets/po03087/irds-skeleton-interpretation.apps_skeletonized_fullskeleton_1500_benign_500spider-natsql-skeleton-context-instruct
Dataset Card for Spider NatSQL Context Instruct
Dataset Summary
Spider is a large-scale complex and cross-domain semantic parsing and text-to-SQL dataset annotated by 11 Yale students
The goal of the Spider challenge is to develop natural language interfaces to cross-domain databases.
This dataset was created to finetune LLMs on the Spider dataset with database context using NatSQL.
NatSQL
NatSQL is an intermediate representation for SQL that simplifies the… See the full description on the dataset page: https://huggingface.co/datasets/richardr1126/spider-natsql-skeleton-context-instruct.preprocessed_hand_skeletonsMulti-VSL-front-skeleton
Multi-VSL (front view) — DWPose skeletons
Whole-body 2D pose keypoints extracted with DWPose
from the front-camera clips of the Multi-VSL Vietnamese Sign Language corpus.
28,406 clips, one .npz per clip
Total size: ~4.6 GB
Laid out as data/<signer>/<clip>.npz — 30 signer directories, 628–1,167 clips
each (HF rejects directories holding more than 10,000 files, so a flat tree is
not possible here)
Contents of each .npz
key
shape
dtype
description
all_xy… See the full description on the dataset page: https://huggingface.co/datasets/Tri1/Multi-VSL-front-skeleton.apps_skeleton_augumented_testapps_skeletonizedskeleton_500_benign_500apps_skeletonized_full-verifier-regressorskeletons_art_prompts
Dataset Card for "skeletons_art_prompts"
More Information needed
customer-support-request-skeletonHandGesture_LandmarkCoordinates_Skeleton
HandGesture_LandmarkCoordinates_Skeleton
Just a datasets for my essay
Very simple with:
Coordinates from 21 Landmark (x0-x20, y0-y20)
8 different labels (thumb_up, thumb_down, point_up, ok_sign, peace, open_hand, nothing, other)
Change Log:
29/08/2025: Uploaded little datasets
17/09/2025: Updated from 2449 rows to 18629 rows
customer-support-requests-skeletoninstagram-character-prompt-skeletonsThis dataset provides prompt skeletons for building
consistent, photorealistic Instagram-style AI characters.
Full visual framework (PDF + images + video):
👉 https://poctavian.gumroad.com/l/cumnev
Instagram Character Prompt Skeletons
This dataset provides clean prompt skeletons designed for building
consistent, photorealistic Instagram-style AI characters.
It focuses on structure, not finished prompts.
What this is
Prompt skeletons (no face/body repetition)
Designed for… See the full description on the dataset page: https://huggingface.co/datasets/Octavian-labs/instagram-character-prompt-skeletons.apps_skeleton_augumented_trainapps_skeletonized_with_harnessesskeleton_slime Disclosure
While its not perfect i hope that you are able to create some nice pieces with it, i am working on improving for the next embedding coming soon, if you have any suggestions or issues please let me know
Usage
To use this embedding you have to download the file and put it into the "\stable-diffusion-webui\embeddings" folder
To use it in a prompt add
art by skeleton slime
add [ ] around it to reduce its weight.
Included Files
6500 steps Usage: art by skeleton slime-… See the full description on the dataset page: https://huggingface.co/datasets/zZWipeoutZz/skeleton_slime.eu-hydro-master-skeleton
EU-Hydro Master Skeleton
Per-basin GeoParquet shards derived from the Copernicus EU-Hydro v1.3 GeoPackages. Four layers are published — river centerlines, river-surface polygons, inland-water polygons (lakes + wide waters), and river-basin polygons — all reprojected to a common CRS and stripped of admin-only columns for easier querying.
Contents
eu_hydro_master_skeleton_geoparquet/
├── river_lines/ # River_Net_l MultiLineString ~1.3 M features
├──… See the full description on the dataset page: https://huggingface.co/datasets/cassini-team-todo/eu-hydro-master-skeleton.