Fitting Room dataset

A dataset of photos taken in fitting rooms  consisting of 100 public images of people standing in front of mirrors or in similar poses, sourced from public websites. The dataset is evenly split between male and female subjects (50 images each), primarily featuring full-body photos that focus on the clothing worn. Each photo includes between two to six clothing items, with a variety of styles represented, though casual attire like t-shirts, jeans, and sneakers are the most common.

The dataset contains a total of 178 men's clothing items and 157 women's clothing items. To test models effectively, the dataset includes images with a wide range of resolutions, both standard and non-standard, as well as varying levels of background complexity. Background complexity is measured by the number of identifiable clothing pieces or other items present in the picture that are not worn by the person. This dataset is intended to evaluate model performance across different metrics.