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
I have a suggestion to make the software lighter
It often happens to me that I use the same image for detection in multiple events
In the current state, I have to create the desired image in each event to detect it
If a special design is made where all images are added and we only connect the images in the events, it would be great
It will both speed up the work and the software does not need to detect a large number of images
Describe the usage
I have a suggestion to make the software lighter
It often happens to me that I use the same image for detection in multiple events
In the current state, I have to create the desired image in each event to detect it
If a special design is made where all images are added and we only connect the images in the events, it would be great
It will both speed up the work and the software does not need to detect a large number of images
Is it a duplicate?
There is no duplicates
The text was updated successfully, but these errors were encountered:
What I mean is that instead of adding images to each event separately, you can select them from a section called Gallery, for example. That Gallery section should also be added to the software and allow you to add images both live and from the phone's gallery.
This will speed up the completion of a scenario, as well as the performance of the software, because it prevents the addition of duplicate images, and as a result, the number of images that are detected will decrease.
Describe the feature
I have a suggestion to make the software lighter
It often happens to me that I use the same image for detection in multiple events
In the current state, I have to create the desired image in each event to detect it
If a special design is made where all images are added and we only connect the images in the events, it would be great
It will both speed up the work and the software does not need to detect a large number of images
Describe the usage
I have a suggestion to make the software lighter
It often happens to me that I use the same image for detection in multiple events
In the current state, I have to create the desired image in each event to detect it
If a special design is made where all images are added and we only connect the images in the events, it would be great
It will both speed up the work and the software does not need to detect a large number of images
Is it a duplicate?
The text was updated successfully, but these errors were encountered: