Exploring Context with Deep Structured Models for Semantic Segmentation
Exploring Context with Deep Structured Models for Semantic Segmentation
复制标题
DOI：
10.1109/tpami.2017.2708714
复制
发表时间：
2018-06-01
影响因子：
23.6
通讯作者：
Reid, Ian
中科院分区：
文献类型：
--
作者：
Lin, Guosheng;Shen, Chunhua;Reid, Ian
我们提出了一种在语义图像分割中利用上下文信息的方法，并特别研究了patch-patch上下文和patch-background上下文在深度cnn中的使用。我们通过结合cnn和条件随机场（CRFs）来建立深度结构化模型，用于学习图像区域之间的patch-patch上下文。具体来说，我们制定了基于cnn的成对势函数来捕获相邻补丁之间的语义相关性。然后对所提出的深度结构模型进行有效的分段训练，以避免在反向传播过程中重复昂贵的CRF推理。对于捕获斑块背景上下文，我们证明了传统的多尺度图像输入和滑动金字塔池的网络设计对于提高性能是非常有效的。我们对所提出的方法进行了综合评价。我们在许多具有挑战性的语义分割数据集上实现了新的最先进的性能。
We propose an approach for exploiting contextual information in semantic image segmentation, and particularly investigate the use of patch-patch context and patch-background context in deep CNNs. We formulate deep structured models by combining CNNs and Conditional Random Fields (CRFs) for learning the patch-patch context between image regions. Specifically, we formulate CNN-based pairwise potential functions to capture semantic correlations between neighboring patches. Efficient piecewise training of the proposed deep structured model is then applied in order to avoid repeated expensive CRF inference during the course of back propagation. For capturing the patch-background context, we show that a network design with traditional multi-scale image inputs and sliding pyramid pooling is very effective for improving performance. We perform comprehensive evaluation of the proposed method. We achieve new state-of-the-art performance on a number of challenging semantic segmentation datasets.