Tiny images

The human visual system is remarkably tolerant to degradations in image resolution: in a scene recognition task, human performance is similar whether32 × 32 color images or multi-mega pixel images are used. With small images, even object recognition and segmentation is per- formed robustly by the visual system, despite the object be- ing unrecognizable in isolation. Motivated by these obser- vations, we explore the space of32 × 32 images using a database of108 32× 32 color images gathered from the In- ternet using image search engines. Each image is loosely labeled with one of the70, 399 non-abstract nouns in En- glish, as listed in the Wordnet lexical database. Hence the image database represents a dense sampling of all object categories and scenes. With this dataset, we use nearest neighbor methods to perform object recognition across the 108 images.