Glossary
Egocentric AI & embodied data glossary
Plain-language definitions of the terms that come up when you evaluate egocentric video for embodied AI. Each entry leads with a one-paragraph answer, then explains why it matters and how we measure it.
- Action segmentA labeled time interval of one sub-action, with start/end timestamps and a verb-noun.
- AnonymizationMaking a person no longer identifiable from data by any reasonably likely means.
- Bimanual manipulationTasks using both hands together, often with coordinated, asymmetric roles.
- Bounding boxThe rectangle of pixel coordinates that marks where a detected object is.
- Consent chainThe documented trail proving every participant granted the rights you need.
- Cross-stream sync errorThe timing offset between simultaneously captured streams like video, depth and IMU.
- Data cardA structured record of how a dataset was collected, consented and licensed.
- De-identificationIrreversibly removing personal identifiers like faces, plates and screens.
- Egocentric videoFirst-person video recorded from a camera worn by the person doing the task.
- Embodied AIAI that perceives and acts through a physical or simulated body.
- Face detectionLocating faces in an image and returning a box and confidence for each.
- Failure-case coverageDeliberately including the hard cases: transit, occlusion, low light, recovery.
- Global-shutter cameraA camera that exposes every pixel simultaneously, avoiding rolling-shutter skew.
- Hand pose estimationRecovering 3D positions of hand and finger joints over time.
- Imitation learningTraining a policy to reproduce demonstrated behavior from example trajectories.
- Instance segmentationPer-pixel masks that identify and separate individual objects over time.
- LeRobot formatAn episodic dataset format used by the LeRobot robot-learning ecosystem.
- PTP clock syncPrecision Time Protocol — sub-microsecond clock alignment across sensors.
- RedactionPermanently obscuring specific regions of an image, video, or document.
- RLDSReinforcement Learning Datasets — an episodic format for robot learning data.
- Skill specA written definition of the exact task, environments and conditions to capture.
- Teleoperation vs egocentric dataRobot-collected demonstrations versus human-worn first-person capture.
- Validated hourAn hour of footage that has passed QA against the spec — not an hour of raw recording.
- WebDatasetA tar-based format for streaming large datasets efficiently during training.