Deep Structured Scene Parsing by Learning with Image Descriptions


This paper addresses a fundamental problem of scene understanding: How to parse the scene image into a structured configuration (i.e., a semantic object hierarchy with object interaction relations) that finely accords with human perception. We propose a deep architecture consisting of two networks: i) a convolutional neural network (CNN) extracting the image representation for pixelwise object labeling and ii) a recursive neural network (RNN) discovering the hierarchical object structure and the inter-object relations. Rather than relying on elaborative user annotations (e.g., manually labeling semantic maps and relations), we train our deep model in a weakly-supervised manner by leveraging the descriptive sentences of the training images. Specifically, we decompose each sentence into a semantic tree consisting of nouns and verb phrases, and facilitate these trees discovering the configurations of the training images. Once these scene configurations are determined, then the parameters of both the CNN and RNN are updated accordingly by back propagation. The entire model training is accomplished through an Expectation-Maximization method. Extensive experiments suggest that our model is capable of producing meaningful and structured scene configurations and achieving more favorable scene labeling performance on PASCAL VOC 2012 over other state-of-the-art weakly-supervised methods.

DOI: 10.1109/CVPR.2016.250

Extracted Key Phrases

9 Figures and Tables

Citations per Year

Citation Velocity: 8

Averaging 8 citations per year over the last 2 years.

Learn more about how we calculate this metric in our FAQ.

Cite this paper

@article{Lin2016DeepSS, title={Deep Structured Scene Parsing by Learning with Image Descriptions}, author={Liang Lin and Guangrun Wang and Rui Zhang and Ruimao Zhang and Xiaodan Liang and Wangmeng Zuo}, journal={2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, year={2016}, pages={2276-2284} }