**Paintings** The paintings were from a larger art collection we created. The full collection contains 5,184 paintings, made up of 18 different paintings by 72 different artists, in each of four different artistic styles (Cubism, Impressionism, Realism, and Renaissance; 288 artists in total). We only used the 1,296 Cubist and 1,296 Impressionist paintings here to equate the number of subcategories within the three image sets (2,592 painting in total). The paintings were collected manually from a variety of websites. We cross referenced the paintings to avoid duplicates, and only downloaded images that were larger than 500×500 pixels. All of the paintings in the art collection were cropped to the centre of the shortest dimension using a 1:1 (square) aspect ratio, and resized (using nearest neighbor scaling) to 256×256 pixels. Any signatures on the paintings were also removed using the “Content Aware” fill tool in Photoshop. Download How-Low-Can-You-Go Paintings collection: https://osf.io/kuja8/ Citation for How-Low-Can-You-Go Paintings collection: Searston, R. A., Thompson, M. B., Vokey, J. R., French, L., & Tangen, J. M. (2018, June 27). How low can you go? Detecting style in extremely low resolution images. https://doi.org/10.17605/OSF.IO/ZX2GS ---------- **Birds** The bird images are a subset of 1,502 extracted from the Cornell Lab of Ornithology’s NABirds V1 collection (downloaded from: http://dl.allaboutbirds.org/nabirds). Seven-hundred and fifty-one images, or half of the subset, are natural photographs of birds from the Accipitridae family (e.g., hawks, eagles, kites, harriers, Old World vultures), and the other half are natural photographs of birds from the Strigidae family (owls). The subset contains sixteen different species, with eight species in each family. The species were randomly selected to match the number of images between the two families, so both sets of eight species contain 120, 103, 103, 99, 73, 72, 95, and 86 images per species respectively. We resized each of the images to 256×256 pixels using a nearest neighbor scaling algorithm and removed any species labels using the “Content Aware” fill tool in Photoshop. Contact Cornell Lab of Ornithology for access to the full original NABirds Dataset: http://dl.allaboutbirds.org/nabirds Download How-Low-Can-You-Go Birds collection: https://osf.io/hd82e/ Citation for How-Low-Can-You-Go Birds collection: Searston, R. A., Thompson, M. B., Vokey, J. R., French, L., & Tangen, J. M. (2018, June 27). How low can you go? Detecting style in extremely low resolution images. https://doi.org/10.17605/OSF.IO/ZX2GS ---------- **Faces** The faces are a collection of 1,772 face images (886 female and 886 male) extracted from the 10k US Adult Faces Database (Bainbridge, Isola, & Oliva, 2013). The entire database contains 10,168 natural face photographs, and we obtained a reduced set of 2,222 images from the database creators along with attribute data (e.g., memorability, attractiveness, happiness, friendliness). We then removed the famous faces from this subset, further reducing it to 2,063 faces (886 female, 1,177 male), and extracted the height and width dimensions of the remaining images. All of the images were 256 pixels tall, except one, which we scaled down to match the height of the others. The width of the images ranged from 153 pixels to 157 pixels, and we resized them to 256×256 pixels. We deleted one image in the male domain that did not match these dimensions. Finally, we randomly sampled 886 male faces from the remaining 1,176, creating an equal number of male and female images. Contact Wilma Bainbridge for access to the full original set of face images and terms of use: https://www.wilmabainbridge.com/facememorability2.html Citation for original 10K US Adult Faces Database: Bainbridge, W.A., Isola, P., & Oliva, A. (2013). The intrinsic memorability of face images. Journal of Experimental Psychology: General, 142(4), 1323-1334. Download How-Low-Can-You-Go Faces collection: https://osf.io/hd82e/ Citation for How-Low-Can-You-Go Faces collection: Searston, R. A., Thompson, M. B., Vokey, J. R., French, L., & Tangen, J. M. (2018, June 27). How low can you go? Detecting style in extremely low resolution images. https://doi.org/10.17605/OSF.IO/ZX2GS