基于区块自适应特征融合的图像实时语义分割 |
| |
引用本文: | 黄庭鸿,聂卓赟,王庆国,李帅,晏来成,郭东生.基于区块自适应特征融合的图像实时语义分割[J].自动化学报,2021,47(5):1137-1148. |
| |
作者姓名: | 黄庭鸿 聂卓赟 王庆国 李帅 晏来成 郭东生 |
| |
作者单位: | 1.华侨大学信息科学与工程学院 厦门 361021 中国 |
| |
基金项目: | 国家自然科学基金61403149华侨大学中青年教师科研提升资助计划项目ZQN-PY408华侨大学中青年教师科研提升资助计划项目Z14Y0002华侨大学研究生科研创新基金17013082039 |
| |
摘 要: | 近年来结合深度学习的图像语义分割方法日益发展, 并在机器人、自动驾驶等领域中得到应用. 本文提出一种基于区块自适应特征融合(Block adaptive feature fusion, BAFF) 的实时语义分割算法, 该算法在轻量卷积网络架构上, 对前后文特征进行分区块自适应加权融合, 有效提高了实时语义分割精度. 首先, 分析卷积网络层间分割特征的感受野对分割结果的影响, 并在跳跃连接结构(SkipNet) 上提出一种特征分区块加权融合机制; 然后, 采用三维卷积进行层间特征整合, 建立基于深度可分离的特征权重计算网络. 最终, 在自适应加权作用下实现区块特征融合. 实验结果表明, 本文算法能够在图像分割的快速性和准确性之间做到很好的平衡, 在复杂场景分割上具有较好的鲁棒性.
|
关 键 词: | 深度学习 实时语义分割网络 区块自适应特征融合 跳跃连接结构 |
收稿时间: | 2018-10-01 |
Real-time Image Semantic Segmentation Based on Block Adaptive Feature Fusion |
| |
Affiliation: | 1.College of Information Science and Engineering, National Huaqiao University, Xiamen 361021, China2.Institute for Intelligent Systems, University of Johannesburg, Johannesburg 2146, South Africa3.the Hong Kong Polytechnic University, Hong Kong 999077, China |
| |
Abstract: | Recently, image semantic segmentation has made great progress with deep learning, which benefits robotics and automatic driving vehicle. This paper proposes a real-time semantic segmentation algorithm based on block adaptive feature fusion (BAFF). Under the framework of a light convolutional network, a block adaptive feature fusion algorithm is proposed in the context-embedding module, to improve the accuracy of real-time semantic segmentation. First, the problem caused by the different size of receptive field in layers is analyzed, and a feature fusion mechanism with block weight is presented on SkipNet. Then, layers' feature integration is carried on by three-dimension convolution. The feature-weights are calculated by an additional network with depthwise-separable-convolutions (DSC). Finally, the features are fused under adaptive weights. Experiments show that this method obtains excellent segmentation results with a good balance between rapidity and accuracy and owns robustness on segmentation of complex scenes. |
| |
Keywords: | |
|
| 点击此处可从《自动化学报》浏览原始摘要信息 |
|
点击此处可从《自动化学报》下载全文 |
|