Size invariance is the ability of an image recognition or classification system to identify an object despite changes in its scale or image size. It is valuable for robust visual perception, although practical systems still have limits when objects become too small or distorted.
In an image classification problem, an algorithm's ability to successfully classify images even when the size of the image changes. For example, the algorithm can still identify a cat whether it consumes 2M pixels or 200K pixels. Note that even the best image classification algorithms still have practical limits on size invariance. For example, an algorithm (or human) is unlikely to correctly classify a cat image consuming only 20 pixels. See also translational invariance and rotational invariance. See the Clustering course for more information.