Episodes Preview yam_bimanual Visualizer
61 episodes · 30 fps · 1 camera · 224×672 av1

DATA-11 streaming test dataset, STACK layout

A small LeRobot v3 dataset for speed tests. It is part of a benchmark of stacking camera views, used by the StreamingLeRobotDataset tests.

Code and results: https://github.com/Hadrien-Cornier/lerobot-camera-stacking-benchmark

Test videos: hadriencornier/lerobot-camera-stacking-benchmark

What is inside

  • 61 episodes, 197,375 frames, 30 fps.
  • Cameras: one video key, observation.images.stack: the 3 views (top, left_wrist, right_wrist) stacked top to bottom in one 224 x 672 frame.
  • 8 video files. Each file is a copy of one STACK group file from the benchmark repo (g000, g001, g002, then again).
  • observation.state and action are random float32 values (14 each). The task text is a placeholder (pick and place). Use this dataset to test reading speed, not to train a policy.
  • Videos: AV1 (libsvtav1), GOP 2, CRF 30, preset 12, yuv420p. The same bytes as the files in the test video repo.

Source

The video frames come from lerobot/abc_130k_v3_train at revision 68651e4929d9fb00f798937b2d62617cab5c771d (Apache-2.0). Credit for the source data goes to the ABC authors: paper and code.

License

Apache-2.0, the same as the source dataset.

Downloads last month
57

Paper for hadriencornier/lerobot-data11-stream-stack