Abstract | ||
---|---|---|
In recent years, skin segmentation has attracted much of attention from computer vision field. Normally, researchers use a simple pre-trained model or define a fixed threshold in color space to deal with skin segmentation. However, it is highly possible to failure in many conditions. In addition, convolutional neural network (CNN) has achieved great success in computer vision. This paper we present a fully convolutional neural network method in skin segmentation. A hand-crafted skin dataset has provided in this study. In the experiment, we attempt many CNN structures to determine the best one. According to the experimental result, we obtained a considerable result in three well-known skin datasets. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/GCCE.2018.8574747 | IEEE Global Conference on Consumer Electronics |
Keywords | Field | DocType |
Convolutional neural network,deep learning,skin segmentation,human skin dataset | Color space,Pattern recognition,Segmentation,Convolutional neural network,Computer science,Human skin,Artificial intelligence,Deep learning | Conference |
ISSN | Citations | PageRank |
2378-8143 | 0 | 0.34 |
References | Authors | |
0 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Chang-Hsian Ma | 1 | 0 | 1.01 |
Huang-Chia Shih | 2 | 187 | 21.98 |