Learning Common and Specific Features for RGB-D Semantic Segmentation with Deconvolutional Networks
This work addresses semantic segmentation for indoor scenes using RGB-D data, presenting an incremental improvement in feature integration for this domain-specific task.
The paper tackles RGB-D semantic segmentation of indoor images by developing a deconvolutional network structure with a feature transformation network that discovers common and specific features between RGB and depth modalities, achieving competitive segmentation accuracy on NYU depth datasets V1 and V2.
In this paper, we tackle the problem of RGB-D semantic segmentation of indoor images. We take advantage of deconvolutional networks which can predict pixel-wise class labels, and develop a new structure for deconvolution of multiple modalities. We propose a novel feature transformation network to bridge the convolutional networks and deconvolutional networks. In the feature transformation network, we correlate the two modalities by discovering common features between them, as well as characterize each modality by discovering modality specific features. With the common features, we not only closely correlate the two modalities, but also allow them to borrow features from each other to enhance the representation of shared information. With specific features, we capture the visual patterns that are only visible in one modality. The proposed network achieves competitive segmentation accuracy on NYU depth dataset V1 and V2.