You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
From the conversation in Gitter, it would be helpful if object detection that produce bounding boxes like words_detect_v2 could somehow be filtered to output the top or best N detections. @beniz suggested this could be done as a chain filter. Creating this issue to track it.
The text was updated successfully, but these errors were encountered:
beniz
changed the title
Single class detection (e.g. words_detect_v2) does not support "best" or "top" N
Object detection (e.g. words_detect_v2) does not support "best" or "top" N
Sep 1, 2020
@cchadowitz-pf thanks, I've updated the scope to object detection at large, i.e. with more than a single class, since this is a demand we hear from several sources now.
I'm also about to revise my strict semantics on the best parameter :)
From the conversation in Gitter, it would be helpful if object detection that produce bounding boxes like
words_detect_v2
could somehow be filtered to output the top or best N detections. @beniz suggested this could be done as a chain filter. Creating this issue to track it.The text was updated successfully, but these errors were encountered: