期刊
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
卷 5, 期 3, 页码 457-467出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TETCI.2019.2947319
关键词
Face; Labeling; Semantics; Image segmentation; Feature extraction; Convolution; Kernel; Detection-segmentation; face labeling; multi-face
资金
- Natural Science Foundation of China [61673187]
- Research Institute in Shenzhen, Huazhong University of Science and Technology, under Shenzhen Innovative Grant [JCYJ20170307160806070]
- NPRP through the Qatar National Research Fund (Qatar Foundation) [NPRP 8-274-2-107]
The paper proposes an end-to-end detection-segmentation system for detailed face labeling, which encodes face images using a pyramid FCN and utilizes three class-specific sub-networks to achieve more accurate single or multi-face labeling results. The method outperforms previous works and achieves state-of-the-art results on the HELEN face dataset.
In this paper, we propose an end-to-end detection-segmentation system to implement detailed face labeling. Fully convolutional networks (FCN) has become the mainstream algorithm in the field of semantic segmentation due to the state-of-the-art performance. However, a general FCN usually produces smooth and homogeneous results. Moreover, when semantic category is extremely unbalanced in samples such as face labeling problem, features for some categories cannot be well explored by FCN. To alleviate these problems, a face image is firstly encoded to multi-level feature maps by a pyramid FCN, then features of different facial components are extracted separately according to the bounding box provided by a one-stage detection head. Three class-specific sub-networks are employed to process the extracted features to obtain the respective segmentation results. The skin-hair region can be decoded directly from the back end of the pyramid FCN. Finally, the overall segmentation result is obtained by combining different branches. Moreover, the proposed method trained on a single-face labeled dataset, can be directly used to implement detailed multi-face labeling tasks without any network modification and additional module or data. The overall structure can be trained in an end-to-end manner while maintaining a small network size (12 MB). Experiments show that the proposed method can generate more accurate (single or multi) face labeling results comparing with previous works and gets the state-of-the-art results in HELEN face dataset.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据