Cross Domain Transfer for Sketch-based Clothing Retrieval

Autor: Guoliang Luo, Fan Yang, Mingwen Wang, Simin Chen, Xiangjian He, Haopeng Lei
Rok vydání: 2020
Předmět:
Zdroj: 2020 8th International Conference on Digital Home (ICDH).
DOI: 10.1109/icdh51081.2020.00012
Popis: Due to the rise of e-commerce platforms, online shopping has become a trend. However, the current mainstream retrieval methods are still limited to using text or exemplar images as input. In the huge commodity database, it remains a long-standing unsolved problem for users to find the interested products quickly. Because the sketch contains more content than the text, and the way to get is more convenient than the exemplar image. In this work, We propose a sketch-based clothing retrieval model. It implements cross-domain retrieval and is used to search for specific clothing. Because most of the existing clothing datasets are only composed of photos, it is difficult to obtain dataset which composed of sketch-photo pairs. Thus, we contribute a clothing dataset, including 34142 sketch-photo pairs. Whether our model in our dataset or other dataset has achieved compelling performance.
Databáze: OpenAIRE