One-Shot Learning
Definition
One-shot learning aims to grasp new tasks based on a single example.
Learning methods that are intended to recognize or adapt a new class or task based on a single labeled example.
Simply explained
A pre-prepared system should recognize a new component type from only one labeled image, for example.
Why it matters in practice
Helps realistically assess model quality, limits, and technical decisions.
Example from business
A system recognizes a new component class after one reference image and existing prior representation.
Limits and pitfalls
- Presenting an example without pretraining, suitable representation, or uncertainty checking as sufficient.
Origin & history
Siamese networks made one-shot learning known for image classification in 2015.
Use in mid-sized companies
- Possible application: A system recognizes a new component class after one reference image and existing prior representation.
Frequently asked questions
What does learning with one example mean?
Learning methods that are intended to recognize or adapt a new class or task based on a single labeled example.
Which mistakes should be avoided?
Presenting an example without pretraining, suitable representation, or uncertainty checking as sufficient.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: