AI Dictionary
One-shot Learning
Definition
An object categorization problem primarily found in computer vision where the model learns from a single example.
Deep Dive
One-shot learning is a sophisticated machine learning paradigm where a model can learn to recognize a new category or object after being exposed to just a single training example of that category. This capability stands in stark contrast to traditional deep learning methods, which typically demand vast datasets comprising thousands or millions of examples to achieve robust performance. It's particularly prevalent in computer vision, addressing challenges where data scarcity is a significant constraint, making it impractical or impossible to collect numerous instances for every potential category.
Examples & Use Cases
- 1A security system learning to recognize a new employee's face from a single photo
- 2Identifying a rare disease from a single pathology slide after training on common diseases
- 3A robotic arm learning to pick up a newly introduced custom-designed part after being shown one instance
Related Terms
Few-shot learningZero-shot learningTransfer learning