Field Effect Deep Networks For Image Recognition With Incomplete Data
Keywords
Deep learning; Image recognition; Incomplete data; Missing features
Abstract
Image recognition with incomplete data is a well-known hard problem in computer vision and machine learning. This article proposes a novel deep learning technique called Field Effect Bilinear Deep Networks (FEBDN) for this problem. To address the difficulties of recognizing incomplete data, we design a novel second-order deep architecture with the Field Effect Restricted Boltzmann Machine, which models the reliability of the delivered information according to the availability of the features. Based on this new architecture, we propose a new three-stage learning procedure with field effect bilinear initialization, field effect abstraction and estimation, and global fine-tuning with missing features adjustment. By integrating the reliability of features into the new learning procedure, the proposed FEBDN can jointly determine the classification boundary and estimate the missing features. FEBDN has demonstrated impressive performance on recognition and estimation tasks in various standard datasets.
Publication Date
8-1-2016
Publication Title
ACM Transactions on Multimedia Computing, Communications and Applications
Volume
12
Issue
4
Document Type
Article
Personal Identifier
scopus
DOI Link
https://doi.org/10.1145/2957754
Copyright Status
Unknown
Socpus ID
84983048011 (Scopus)
Source API URL
https://api.elsevier.com/content/abstract/scopus_id/84983048011
STARS Citation
Zhong, Sheng Hua; Liu, Yan; and Hu, Kien A., "Field Effect Deep Networks For Image Recognition With Incomplete Data" (2016). Scopus Export 2015-2019. 2319.
https://stars.library.ucf.edu/scopus2015/2319