Autotag Training Set Queries
Autotag Training Set Query - Examples
“Show me all of the images I manually selected as positive examples while creating and refining my ‘rain’ Autotag”
autotag_training.image.rain = positive
“Show me all of the images I specifically excluded from my ‘speed_limit’ Object Autotag while refining it”
autotag_training.object.speed_limit = manual_negative
“Show me all of the images that my ‘rain’ Autotag found that I didn’t manually select as a positive example during Autotag creation”
autotag_training.image.rain != positive AND autotag.image.rain > 0
Autotag Training Set Query - Details
Sometimes it is useful to query for the specific images or objects that you labeled (or skipped during refinement, which is treated as an explicit “negative” label) while creating an Autotag.
autotag_training.image.[tag_name] | If the image was used during Autotag creation or refinement, will be positive if the image was selected and manual_negative if the image was skipped during refinement. |
autotag_training.object.[tag_name] | If the object was used during Autotag creation or refinement, will be positive if the image was selected and manual_negative if the image was skipped during refinement. |
Updated almost 3 years ago