Skip to content
Discussion options

You must be logged in to vote

You can decouple the bottleneck by using ObserveOn. This way the videowriter/acquisition should not have to wait on the thread from the pose estimation.

Alternatively (which is a simpler solution) simply use a Slice operator to downsample the input to the inference operator.

Replies: 1 comment 1 reply

Comment options

You must be logged in to vote
1 reply
@edmundchong
Comment options

Answer selected by edmundchong
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants