Please use this identifier to cite or link to this item:
Title: Structure-aware multimodal feature fusion for RGB-D scene classification and beyond
Authors: Wang, Anran
Cai, Jianfei
Lu, Jiwen
Cham, Tat-Jen
Keywords: Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
Issue Date: 2018
Source: Wang, A., Cai, J., Lu, J., & Cham, T.-J. (2018). Structure-aware multimodal feature fusion for RGB-D scene classification and beyond. ACM Transactions on Multimedia Computing, Communications, and Applications, 14(2s), 39-. doi:10.1145/3115932
Journal: ACM Transactions on Multimedia Computing, Communications, and Applications
Abstract: While convolutional neural networks (CNNs) have been excellent for object recognition, the greater spatial variability in scene images typically means that the standard full-image CNN features are suboptimal for scene classification. In this article, we investigate a framework allowing greater spatial flexibility, in which the Fisher vector (FV)-encoded distribution of local CNN features, obtained from a multitude of region proposals per image, is considered instead. The CNN features are computed from an augmented pixel-wise representation consisting of multiple modalities of RGB, HHA, and surface normals, as extracted from RGB-D data. More significantly, we make two postulates: (1) component sparsity—that only a small variety of region proposals and their corresponding FV GMM components contribute to scene discriminability, and (2) modal nonsparsity—that features from all modalities are encouraged to coexist. In our proposed feature fusion framework, these are implemented through regularization terms that apply group lasso to GMM components and exclusive group lasso across modalities. By learning and combining regressors for both proposal-based FV features and global CNN features, we are able to achieve state-of-the-art scene classification performance on the SUNRGBD Dataset and NYU Depth Dataset V2. Moreover, we further apply our feature fusion framework on an action recognition task to demonstrate that our framework can be generalized for other multimodal well-structured features. In particular, for action recognition, we enforce interpart sparsity to choose more discriminative body parts, and intermodal nonsparsity to make informative features from both appearance and motion modalities coexist. Experimental results on the JHMDB and MPII Cooking Datasets show that our feature fusion is also very effective for action recognition, achieving very competitive performance compared with the state of the art.
ISSN: 1551-6857
DOI: 10.1145/3115932
Rights: © 2018 Association for Computing Machinery (ACM). All rights reserved.
Fulltext Permission: none
Fulltext Availability: No Fulltext
Appears in Collections:IMI Journal Articles

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.