A Clothes Classification Method Based on the gcForest
Li Han, Haihong Zhao, Erxin Yu, Bai Yuming, Huiying Li · 2018
In this paper, we propose a clothes classification method, an application of the gcForest. The gcForest, a novel decision tree ensemble method proposed by Zhou and Feng, performs excellently in many experiments. Moreover, we come up with some adjustment approaches for gcForest. Using these adjustment approaches, we get a satisfactory result when gcForest is used for clothes classification task. Furthermore, we create a data set for clothes classification which contains 8 categories of clothes including a total of 52000 images. And we preprocess the images in the data set by GrabCut-based clothes extraction algorithm.