- Description :
Ensemble de données réel d'un seul bras robot démontrant 12 compétences de manipulation non triviales sur 38 tâches, 7 500 trajectoires.
Page d'accueil : https://robopen.github.io/
Code source :
tfds.robotics.rtx.RoboSetVersions :
-
0.1.0(par défaut) : version initiale.
-
Taille du téléchargement :
Unknown sizeTaille de l'ensemble de données :
179.42 GiBMise en cache automatique ( documentation ) : Non
Divisions :
| Diviser | Exemples |
|---|---|
'train' | 18 250 |
- Structure des fonctionnalités :
FeaturesDict({
'episode_metadata': FeaturesDict({
'file_path': string,
'trial_id': string,
}),
'steps': Dataset({
'action': Tensor(shape=(8,), dtype=float32),
'discount': Scalar(shape=(), dtype=float32),
'is_first': bool,
'is_last': bool,
'is_terminal': bool,
'language_instruction': string,
'observation': FeaturesDict({
'image_left': Image(shape=(240, 424, 3), dtype=uint8),
'image_right': Image(shape=(240, 424, 3), dtype=uint8),
'image_top': Image(shape=(240, 424, 3), dtype=uint8),
'image_wrist': Image(shape=(240, 424, 3), dtype=uint8),
'state': Tensor(shape=(8,), dtype=float32),
'state_velocity': Tensor(shape=(8,), dtype=float32),
}),
'reward': Scalar(shape=(), dtype=float32),
}),
})
- Documentation des fonctionnalités :
| Fonctionnalité | Classe | Forme | Type D | Description |
|---|---|---|---|---|
| FonctionnalitésDict | ||||
| épisode_métadonnées | FonctionnalitésDict | |||
| épisode_metadata/file_path | Tenseur | chaîne | ||
| épisode_metadata/trial_id | Tenseur | chaîne | ||
| mesures | Ensemble de données | |||
| étapes/actions | Tenseur | (8,) | flotteur32 | |
| étapes/remise | Scalaire | flotteur32 | ||
| étapes/is_first | Tenseur | bouffon | ||
| étapes/est_dernier | Tenseur | bouffon | ||
| étapes/is_terminal | Tenseur | bouffon | ||
| étapes/instruction_langue | Tenseur | chaîne | ||
| étapes/observation | FonctionnalitésDict | |||
| étapes/observation/image_left | Image | (240, 424, 3) | uint8 | |
| étapes/observation/image_right | Image | (240, 424, 3) | uint8 | |
| étapes/observation/image_top | Image | (240, 424, 3) | uint8 | |
| étapes/observation/image_poignet | Image | (240, 424, 3) | uint8 | |
| étapes/observation/état | Tenseur | (8,) | flotteur32 | |
| étapes/observation/state_velocity | Tenseur | (8,) | flotteur32 | |
| étapes/récompense | Scalaire | flotteur32 |
Clés supervisées (Voir doc
as_supervised) :NoneFigure ( tfds.show_examples ) : non pris en charge.
Exemples ( tfds.as_dataframe ) :
- Citation :
@misc{bharadhwaj2023roboagent, title={RoboAgent: Generalization and Efficiency in Robot Manipulation via Semantic Augmentations and Action Chunking}, author={Homanga Bharadhwaj and Jay Vakil and Mohit Sharma and Abhinav Gupta and Shubham Tulsiani and Vikash Kumar}, year={2023}, eprint={2309.01918}, archivePrefix={arXiv}, primaryClass={cs.RO} }