Street Fashion Style (SFS) dataset

The Street Fashion Style (SFS) dataset is a new street photos dataset collected from Chictopia, where a total of 293,105 user posts are crawled. In each post, a user usually publishes the photograph of her/his worn outfit along with associated tags. Generally, these tags include current season, the suitable occasion, fashion style, the detailed garment information (e.g. category, color and brand), the geographical and year information.

The dataset can be applied, but not limited to the following research areas:

• multi-task learning
• feature embedding learning
• fashion-related classification
• fashion trends analysis

Please cite the following paper if you use the SFS dataset in your work (papers, articles, reports, books, software, etc):

• X. Gu, Y. Wong, P. Peng, L. Shou, G. Chen, M. Kankanhalli
  Understanding Fashion Trends from Street Photos via Neighbor-Constrained Embedding Learning
  ACM Multimedia, 2017.
  http://doi.org/10.1145/3123266.3123441

After downloading all parts (images-*), extract using: cat images-* | tar zx

Alternate download mirror - https://pan.baidu.com/s/1nvA1IPz