Where to start
Inspect the dataset structure and metadata before modifying or combining episodes.
Use it in your project
Write a schema audit for one episode, including observations, actions, time and provenance.
Scope and limitations
Dataset versions differ. Do not assume that examples for one format apply unchanged to another.
Follow a related roadmap
Collect demonstrations and train an imitation policy · Beginner →
Read: Inspect the dataset structure and metadata before modifying or combining episodes.
Use the data structure to make episodes, observations, actions and processing provenance inspectable.
Produce: Audit one episode from camera observations through action/state metadata and time alignment.
Build a perception dataset and evaluation study · Professional →
Read: Inspect the dataset structure and metadata before modifying or combining episodes.
Use the data structure to make episodes, observations, actions and processing provenance inspectable.
Produce: Publish data organization, versions and provenance so another researcher can reconstruct an episode.
Plan a shared robotics research lab · Intermediate →
Read: Inspect the dataset structure and metadata before modifying or combining episodes.
Use the data structure to make episodes, observations, actions and processing provenance inspectable.
Produce: Design a shared experiment/data ledger that records platform, calibration and software identity.