Simple Shapes Dataset

This dataset is used in the paper Semi-supervised Multimodal Representation Learning through a Global Workspace, Devillers et al., 2023 (under review). To use this dataset, use the code provided here: https://github.com/bdvllrs/bimGW. It consists of 32x32 pixel images of shapes with multiple attributes (size, location, rotation, color). Each image is also paired with its ground truth information (attributes), and a natural language description (English) of the image. The dataset is composed of: a train set of 500,000 samples, a val and a test set of 1000 samples each. It also contains already processed 12-dimensional visual features (from a VAE), and presaved BERT features of the text descriptions