Context Embedding Networks

Kun Ho Kim, Oisin Mac Aodha, Pietro Perona

Low dimensional embeddings that capture the main variations of interest in collections of data are important for many applications. One way to construct these embeddings is to acqu…