Stereoscopic image retargeting based on 3D saliency detection

Wang, Junle; Fang, Yuming; Narwaria, Manish; Wang, Lin; Callet, Patrick Le

doi:10.1109/icassp.2014.6853680

Cited by 14 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Their stereo retargeting algorithm achieves state-of-the-art performance in terms of their depth distortion metric. Very recently, Wang et al [18] propose to take viewing bias into account to seam carve stereo images. Nevertheless, these methods strongly rely on accurate disparity maps to first construct an energy function involving two views and secondly to find the seam pixels in a view given those in the other view.…”

Section: Related Workmentioning

confidence: 99%

“…Because of its elegance and simplicity, several seam-based retargeting algorithms have been developed in recent years including notable stereo extensions such as [15,3,18]. However, inheriting the discrete nature of seam carving, stereo seam carving algorithms have been criticized for their relatively weak structure protection abilities as compared with continuous retargeting approaches.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

On Preserving Structure in Stereo Seam Carving

Lien

Turk

2015

2015 International Conference on 3D Vision

View full text Add to dashboard Cite

The major objective of image retargeting algorithms is to preserve the viewer's perception while adjusting the aspect ratio of an image. This means that an ideal retargeting algorithm has to be able to preserve high-level semantics and avoid generating low-level image distortion. Stereoscopic image retargeting poses a even more challenging problem in that the 3D perception has to be preserved as well. In this paper, we propose an algorithm based on high-order twoview co-labeling to simultaneously retarget a given stereo pair and preserve its 2D as well as 3D quality. Our experimental results qualitatively demonstrate the improved ability of preserving 2D image structures in both views. In addition, we show quantitatively that our algorithm improves upon the state-of-the-art up to 85% in terms of a measurement based on depth distortion.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

On Preserving Structure in Stereo Seam Carving

Lien

Turk

2015

2015 International Conference on 3D Vision

View full text Add to dashboard Cite

show abstract

“…Correspondingly, the optimal stereo couple seams can be selected and removed with fusing the 3D Fig. 6 The technical block diagram of 3D video retargeting saliency map, energy map and visibility of pixels by dynamic programming [32].…”

Section: Content-aware 3d Video Retargetingmentioning

confidence: 99%

3D Video Rendering Adaptation: A Survey

Liu

et al. 2015

3D Res

View full text Add to dashboard Cite

With various types of 3D display, the same 3D content may present different visual experience qualities. To achieve the perfect 3D visual experience, 3D video rendering adaptation is highly desired. This paper surveys the state-of-the-art 3D video rendering adaption technologies and further summarizes the remaining open challenges. First, we explain the basic 3D vision principle which provides 3D video rendering adaptation rules. Second, based on the outline of disparity processing, we survey two categories of 3D video rendering adaptation technologies: (1) video size and depth range adaptation; (2) viewpoint adaptation. For video size and depth range adaptation, the linear scaling based 3D video super-resolution, disparity transform and content-aware 3D video retargeting are specifically overviewed. Finally, we point out the open issues to be investigated in the area of 3D video rendering adaptation. Graphical AbstractKeywords 3D video rendering adaptation Á 3D video resize Á Disparity transform Á 3D video retargeting Á Viewpoint adaptation 1 Introduction

show abstract

“…According to the human visual attention mechanism (HVAM), this task imitates the human visual system through a computer to extract the interesting/remarkable regions that naturally draw human visual attention in a scene. It helps to advance the typical computer vision studies, such as image segmentation [2], content and video compression [3,4], object tracking and retargeting [5,6], object recognition [7], etc.…”

Section: Introductionmentioning

confidence: 99%

A joint similarity matrix learning of multi-view data for RGB-T saliency detection

Wang,

She,

Ding

et al. 2024

Preprint

View full text Add to dashboard Cite

Graph-based saliency detection has aroused considerable attention due to its ability to extract object of interest from the natural scenes. However, the most existing methods usually consisted the graph constrcution and initial back-gorund/foreground seeds selection, which may simplify the relationship among multi-view data and results in an inaccurate in complicated scenes. To our knowledge , the success of the graph-based saliency detection methods mainly rely on the qulaity of graph. In this paper, we target to achieve a joint similarity matrix learning (AJSML) from multi-view data based on graph diffusion process, which is committed to facilitating the RGB-T saliency detection task. Our assumption is that salient object in the complicaed scene always tend to be similarity appearance and compactness distribution in spatial domain. Specifically, we first design a generalized framework to simultaneously learn the high quality graph and similarity matrix for multi-view data. Thus, the similarity relationship and correlation information of multi-view data can be effectively diffused on the high quality graph, which is conducive to generating a faithful similarity matrix. Further, we present a post-processing technology called an adaptive weighted semi-supervised 1 learning (AWSL), integrating saliency information and cross-modality graphs, is developed to promote the accuracy degree of saliency results. Finally, extensive experimental results on well-known benchmark RGB-T, RGB, and RGB-D datasets demonstrate the superiority of the proposed method, in comparison to several state-of-the-art methods.

show abstract

Stereoscopic image retargeting based on 3D saliency detection

Cited by 14 publications

References 11 publications

On Preserving Structure in Stereo Seam Carving

On Preserving Structure in Stereo Seam Carving

3D Video Rendering Adaptation: A Survey

A joint similarity matrix learning of multi-view data for RGB-T saliency detection

Contact Info

Product

Resources

About