Datasets:
How does key-frame annotation for each sub-step generate？
#29
by
q9177
- opened
Your paper said "We also provide key-frame and instruction annotations for each sub-step to facilitate policy learning in such challenging scenarios."
Question:
May I ask how these key-frames and textual annotations for sub-steps are generated? Are they manually segmented and labeled, or are they automatically generated by algorithms? If algorithms are used, which specific models are employed for automatic segmentation and annotation? Thanks very much!