DeepFashion: Powering Robust Clothes Recognition and Retrieval with Rich Annotations

TLDR

Recent advances in clothes recognition have been driven by the construction of clothes datasets, yet existing datasets lack extensive annotations and struggle with real‑world challenges. This paper introduces DeepFashion1, a large‑scale clothes dataset with comprehensive annotations, and proposes FashionNet, a deep model that jointly predicts clothing attributes and landmarks. DeepFashion1 comprises over 800,000 images richly annotated with attributes, landmarks, and cross‑scenario correspondences, while FashionNet iteratively optimizes feature learning by using predicted landmarks to pool or gate features. The rich annotations of DeepFashion enable powerful clothes‑recognition algorithms, and experiments show that FashionNet effectively leverages them, demonstrating the dataset’s usefulness.

Abstract

Recent advances in clothes recognition have been driven by the construction of clothes datasets. Existing datasets are limited in the amount of annotations and are difficult to cope with the various challenges in real-world applications. In this work, we introduce DeepFashion1, a large-scale clothes dataset with comprehensive annotations. It contains over 800,000 images, which are richly annotated with massive attributes, clothing landmarks, and correspondence of images taken under different scenarios including store, street snapshot, and consumer. Such rich annotations enable the development of powerful algorithms in clothes recognition and facilitating future researches. To demonstrate the advantages of DeepFashion, we propose a new deep model, namely FashionNet, which learns clothing features by jointly predicting clothing attributes and landmarks. The estimated landmarks are then employed to pool or gate the learned features. It is optimized in an iterative manner. Extensive experiments demonstrate the effectiveness of FashionNet and the usefulness of DeepFashion.