How to perform sequential inference instead of whole video process?
jiangzhengkai opened this issue · 1 comments
jiangzhengkai commented
Thanks for your great work. I want to know how to perform sequential inference instead of the whole video process.
sukjunhwang commented
Hi @jiangzhengkai,
cfg.INPUT.SAMPLING_FRAME_NUM is what controls the input clip length.
As the parameter is set to 5 by default, it would run in a sequential inference mode unless modified.