Glossary
Action segment
A labeled time interval of one sub-action, with start/end timestamps and a verb-noun.
Definition
An action segment is a labeled time interval marking one sub-action within an episode — a start and end timestamp with a verb and noun, such as “grasp mug”. It turns a continuous clip into supervised, temporally-localized units a policy can learn from.
01
Why it matters
Frame-accurate action boundaries let you train and evaluate on the moments that matter and skip the transit. Narrations or clip-level tags are not enough for manipulation, where the exact contact moment is the signal.
02
How Firsthand handles it
Firsthand ships action_segments as {start_ns, end_ns, verb, noun} with nanosecond timestamps aligned to every other stream.
Related terms