Segmenting Robot Video into Actionable Subtasks
A benchmark and field report on using VLMs to turn robot and egocentric video into timestamped subtask annotations.
Summary not yet generated for this article.
Comments (2 of ) ⌛