Glossary term
Glossary term
Multimodal AI
In an image classification problem, an algorithm's ability to successfully classify images even when the orientation of the image changes. For example, the algorithm can still identify a tennis racket whether it is pointing up, sideways, or down. Note that rotational invariance is not always desirable; for example, an upside-down 9 shouldn't be classified as a 9.
See also translational invariance and size invariance.
Created for this library
A medical imaging team augments training images with rotations so its classifier achieves rotational invariance on slides.
A robotics perception team trains its object detector with random rotations so identification is robust across object orientations.
A retail analytics team augments product images with rotations so its detector recognizes products at any orientation on shelves.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License