What does [1, 512] tensor of features represent? #899
Unanswered
tejask-git
asked this question in
Q&A
Replies: 1 comment 1 reply
-
They are embeddings. A learned representation of appearances, where appearances that are similar, have similar representations. |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
The method _get_features returns a [1, 512] tensor of the detected object's cropped, resized image - https://github.com/mikel-brostrom/yolov8_tracking/blob/737b6ae62b206f718781d3de2398859010e6dcd9/boxmot/strongsort/strong_sort.py#L53
Is it used to make sure that the detected object doesn't move rapidly, or goes out of frame, etc. while it is being tracked?
What does that [1, 512] tensor represent?
Beta Was this translation helpful? Give feedback.
All reactions