Locally isometric and conformal parameterization of image manifold

2015 ◽  
Author(s):  
A. V. Bernstein ◽  
A. P. Kuleshov ◽  
Yu. A. Yanovich
Keyword(s):  
Image Manifold

scholarly journals Generative Visual Manipulation on the Natural Image Manifold

2016 ◽  
pp. 597-613 ◽  
Author(s):  
Jun-Yan Zhu ◽  
Philipp Krähenbühl ◽  
Eli Shechtman ◽  
Alexei A. Efros
Keyword(s):  
Natural Image ◽  
Image Manifold

Dynamic Building Tracking from UAVs Based on Image Manifold Learning

2013 ◽  
Vol 756-759 ◽  
pp. 4121-4125
Author(s):  
Peng Zhang ◽  
Yuan Yuan Ren
Keyword(s):  
Unmanned Aerial Vehicles ◽  
Visual Tracking ◽  
Target Recognition ◽  
Perceptual Information ◽  
System Control ◽  
Aerial Vehicles ◽  
Robust Feature Extraction ◽  
Image Manifold ◽  
Low Dimensional ◽  
Low Dimensional Features

Fast and accurate visual tracking of ground buildings can provide unmanned aerial vehicles (UAVs) with rich perceptual information, which is very important for target recognition, navigation and system control. However, when an UAV moves fast, both background and buildings in visual scenes change relatively and rapidly. Consequently, there are no constant features for objects' appearance, which poses great challenges for visual tracking of buildings. In this paper, we first build an image manifold of buildings, which can encode the continuous variation of appearance. We then propose an efficient approach to learn this manifold and obtain more robust feature extraction results. By using a simple tracking framework, we successfully apply the extracted low-dimensional features to real-time building tracking. Experimental results demonstrate the effectiveness of the proposed method.

A New Extension of Locality Preserving Projections for Image Feature Extraction

2012 ◽  
Vol 241-244 ◽  
pp. 1715-1718
Author(s):  
Guo Hong Huang
Keyword(s):  
Feature Extraction ◽  
Local Information ◽  
Image Feature ◽  
Two Dimensional ◽  
Two Stage ◽  
Image Feature Extraction ◽  
Locality Preserving ◽  
Image Manifold ◽  
New Framework ◽  
Novel Algorithm

This paper proposes a novel algorithm for image feature extraction, namely, the two-directional two-dimensional locality preserving projection, ((2D)2LPP), which can find an embedding from two directions that not only preserves local information and detect the intrinsic image manifold structure, but also combines the both information between rows and those between columns simultaneously. We also combine the advantages of (2D)2LPP and LDA, and propose a new framework for feature extraction as two-stage: “(2D)2LPP+LDA.” The LDA step is performed to further reduce the dimension of feature matrix in the (2D)2LPP subspace. Experimental results on ORL face databases demonstrate the effectiveness of the proposed methods.

Learning image manifold via local tensor subspace alignment

2014 ◽  
Vol 139 ◽  
pp. 22-33 ◽  
Author(s):  
Songsong Wu ◽  
Xiaoyuan Jing ◽  
Zhisen Wei ◽  
Jian Yang ◽  
Jingyu Yang
Keyword(s):  
Image Manifold ◽  
Tensor Subspace ◽  
Subspace Alignment

Learning a hierarchical image manifold for Web image classification

2012 ◽  
Vol 13 (10) ◽  
pp. 719-735
Author(s):  
Rong Zhu ◽  
Min Yao ◽  
Li-hua Ye ◽  
Jun-ying Xuan
Keyword(s):  
Image Classification ◽  
Image Manifold

scholarly journals Human Sensitivity to Perturbations Constrained by a Model of the Natural Image Manifold

2018 ◽  
Author(s):  
Ingo Fruend ◽  
Elee Stalker
Keyword(s):  
Image Segmentation ◽  
Statistical Properties ◽  
Natural Images ◽  
Generative Adversarial Networks ◽  
Natural Image ◽  
Quantitative Characterization ◽  
Adversarial Networks ◽  
Image Manifold ◽  
Visual Structure

Humans are remarkably well tuned to the statistical properties of natural images. However, quantitative characterization of processing within the domain of natural images has been difficult because most parametric manipulations of a natural image make that image appear less natural. We used generative adversarial networks (GANs) to constrain parametric manipulations to remain within an approximation of the manifold of natural images. In the first experiment, 7 observers decided which one of two synthetic perturbed images matched a synthetic unperturbed comparison image. Observers were significantly more sensitive to perturbations that were constrained to an approximate manifold of natural images than they were to perturbations applied directly in pixel space. Trial by trial errors were consistent with the idea that these perturbations disrupt configural aspects of visual structure used in image segmentation. In a second experiment, 5 observers discriminated paths along the image manifold as recovered by the GAN. Observers were remarkably good at this task, confirming that observers were tuned to fairly detailed properties of an approximate manifold of natural images. We conclude that human tuning to natural images is more general than detecting deviations from natural appearance, and that humans have, to some extent, access to detailed interrelations between natural images.

Sign in / Sign up

Export Citation Format

Share Document