Examples of our coarse annotations can be found here. These annotations are intended to support research that aims to exploit large volumes of (weakly) annotated data, e.g. for training deep neural networks.
Monthly Archives: August 2015