BROWSE
Detail
Conference Paper
Hair Segmentation Using Deep Mobile Optimization
Cited - time in 
Share 
Share
- Authors
- SeongWoo Park, ChanWoong Kwak, Ho-Sub Yoon
- Issue Date
- 2019-06
- Citation
- International Conference on Ubiquitous Robots (UR) 2019, pp.1-2
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- In this paper, we deal with the problem about hair segmentation in static images. We propose a deep learning method. Among them, we use one of the deep learning models, named MobileNetV2, which is applied to U-net. We will define it as MobileUnet. The training image we used was about 6000 images, and the data augmentation further increased the volume about 14 times. This proposed network was evaluated on some of CelebA databases. The experiment results, we performed with the hair segmentation rate about 95% demonstrate why deep learning has efficiency about the purposed solution. This paper is currently in progress and it has not yet been fully completed.
- KSP Keywords
- Data Augmentation, Deep learning method, Experiment results, Training image, deep learning(DL), deep learning models, hair segmentation