Hand-designed features such as LBP, SIFT and HOG have gained great success during the past years in various vision tasks such as object recognition, target tracking and action recognition. However, these only capture low-level textural or edge information and it has proven difficult to design features that effectively capture mid-level cues (e.g. edge intersections) or high-level representation (e.g. object parts). Recently, developments in machine learning, known as "Deep Learning", have shown how hierarchies of features can be learned in an unsupervised manner directly from data. The learned features are capable of extracting mid-level or high-level information. More importantly, these features are learned from the data, which are more effective for the specific task at hand. Some preliminary experimental results in various tasks have validated their powerful ability, which attracts increasing interest in the computer vision community.

Very recently, a huge number of efforts have being devoted to develop novel and more effective unsupervised feature learning methods. The primary purpose of this special issue is to organize a collection of recently developed unsupervised feature learning models for various vision tasks. The special issue is intended to be an international forum for researchers to report the recent developments in this field in an original research paper style. The topics include, but are not limited to:

· Energy-based models and their fast optimization solutions

· Form and motion factorization from video data

· Invariance feature learning

· Semantic feature learning

· Deep learning model on large-scale data

· Visualization of the learned features

· Multi-modal feature learning

· Real-time vision applications

Important Dates

Manuscript Due: August 28, 2014

First Round of Reviews: September 30, 2014

Second round of Reviews: December 15, 2014

Notification of acceptance: January 20, 2015

Publication Date: June 2015

Submission Guidelines

Prospective authors should prepare and submit their manuscripts according to the IET Computer Vision author guidelines (http://digital-library.theiet.org/journals/author-guide).

Guest Editors

Shengping Zhang

Brown University, United States

shengping_zhang@brown.edu

Baochang Zhang

Beihang University, China

bczhang@buaa.edu.cn

Qixiang Ye

University of Maryland, United States

qxye@umiacs.umd.edu