Model Training & Adaptation
Transfer learning
Reusing knowledge learned for one task to improve performance on another.
Example
A vision model trained on general images is adapted for detecting defects in factory parts.
Why it matters
It can save time and examples by building on abilities already learned elsewhere.
You'll hear it
“Can we adapt what it already recognizes for this new task?”
Can you control it?
Set by the team
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
- Can unrelated earlier learning hurt performance?
- Yes. Unhelpful patterns can make the new task harder rather than easier.
- Does transfer mean no new learning is needed?
- Not necessarily. Adaptation and task-specific examples may still be required.
- Can it transfer between very different kinds of tasks?
- Sometimes, but usefulness depends on what patterns the tasks genuinely share.