Images of 160 pairs of athletic shoes in two styles and four sizes. Participants wore the shoes for at least 10,000 steps per week, and images of the shoes were taken at five-week intervals, producing five observation points for each pair of shoes.
A database of two-dimensional images of footwear outsole scans. There are 1,500 impressions from 150 pairs of used shoes. The information about the database contains the ID of the users, genders (for size), brands, models and sizes.
A database of impressions from 330 shoes: 165 made in blood and 165 made in dust. A different subset of shoes was used for each impression type, and there is an unenhanced and an enhanced version of each impression. All impressions are oriented with the toe at the top and the heel at the bottom.
ShoeRinsics learns to predict shoe-tread depth from fully supervised synthetic data and unsupervised retail image data. This database includes two training data sets and two validation data sets consisting of shoe-tread images and print pairs.