Learning Type-Aware Embeddings for Fashion Compatibility
This work addresses the challenge of building coherent outfits in online fashion for users and platforms, though it is incremental as it builds on existing embedding methods with type-aware modifications.
The paper tackled the problem of learning embeddings for fashion items that capture both similarity and compatibility across different types, achieving a 3-5% improvement over state-of-the-art methods on outfit compatibility prediction and fill-in-the-blank tasks using datasets from Polyvore.
Outfits in online fashion data are composed of items of many different types (e.g. top, bottom, shoes) that share some stylistic relationship with one another. A representation for building outfits requires a method that can learn both notions of similarity (for example, when two tops are interchangeable) and compatibility (items of possibly different type that can go together in an outfit). This paper presents an approach to learning an image embedding that respects item type, and jointly learns notions of item similarity and compatibility in an end-to-end model. To evaluate the learned representation, we crawled 68,306 outfits created by users on the Polyvore website. Our approach obtains 3-5% improvement over the state-of-the-art on outfit compatibility prediction and fill-in-the-blank tasks using our dataset, as well as an established smaller dataset, while supporting a variety of useful queries.