Model Training & Adaptation
Multilabel classification
Allowing something to belong to several groups at once.
Example
A photo receives both 'outdoors' and 'bicycle' labels.
Why people use it
It handles cases that belong to several groups at once.
What you'll hear
“This photo can be both outdoors and a family picture.”
What this means for you
Check which labels are missed or wrongly added, rather than only the overall score.
Can you control it?
Developer-only
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Must the labels exclude one another?
- No. Multiple labels can be true at the same time.
- Can it miss one label while finding another correctly?
- Yes. A partly correct set of labels can still leave out something important.
- Does every label need the same confidence cutoff?
- No. Different labels can use different rules for when they should be applied.