English

DAG-Recurrent Neural Networks For Scene Labeling

Computer Vision and Pattern Recognition 2015-11-24 v2

Abstract

In image labeling, local representations for image units are usually generated from their surrounding image patches, thus long-range contextual information is not effectively encoded. In this paper, we introduce recurrent neural networks (RNNs) to address this issue. Specifically, directed acyclic graph RNNs (DAG-RNNs) are proposed to process DAG-structured images, which enables the network to model long-range semantic dependencies among image units. Our DAG-RNNs are capable of tremendously enhancing the discriminative power of local representations, which significantly benefits the local classification. Meanwhile, we propose a novel class weighting function that attends to rare classes, which phenomenally boosts the recognition accuracy for non-frequent classes. Integrating with convolution and deconvolution layers, our DAG-RNNs achieve new state-of-the-art results on the challenging SiftFlow, CamVid and Barcelona benchmarks.

Keywords

Cite

@article{arxiv.1509.00552,
  title  = {DAG-Recurrent Neural Networks For Scene Labeling},
  author = {Bing Shuai and Zhen Zuo and Gang Wang and Bing Wang},
  journal= {arXiv preprint arXiv:1509.00552},
  year   = {2015}
}
R2 v1 2026-06-22T10:47:05.835Z