Add multithreading #186

Closed
opened 2024-07-17 18:23:05 +00:00 by jgeorgi · 2 comments
Owner

The next level of performance improvement is most definitely multithreading the detection system (especially for the live camera view) using web workers.

The next level of performance improvement is most definitely multithreading the detection system (especially for the live camera view) using web workers.
jgeorgi added this to the 0.5.0 milestone 2024-07-17 18:23:05 +00:00
jgeorgi added the
enhancement
label 2024-07-17 18:23:05 +00:00
Author
Owner

This requires a fairly large rebuild. The image as background for the canvas is a problem because getting the image data to the worker in a way that tensorflow can handle is not easy. So, probably adding to the image to the canvas context and then allowing the canvas object to be passed to the worker is probably the best bet.

This requires a fairly large rebuild. The image as background for the canvas is a problem because getting the image data to the worker in a way that tensorflow can handle is not easy. So, probably adding to the image to the canvas context and then allowing the canvas object to be passed to the worker is probably the best bet.
Author
Owner

Nevermind, figured out how to pass image data more efficiently, so the old background image technique still works. Don't need to move the image to the canvas.

Until I want panning and zooming, anyway...

Nevermind, figured out how to pass image data more efficiently, so the old background image technique still works. Don't need to move the image to the canvas. Until I want panning and zooming, anyway...
jgeorgi referenced this issue from a commit 2024-07-25 17:56:22 +00:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: ALVINN/ALVINN_f7#186
No description provided.