Learning-based depth estimation from 2D images using GIST and saliency

Arciniegas Herrera, Jose Luis and Konrad, Janusz and Blanco Adán, Carlos Roberto del and García Santos, Narciso (2015). Learning-based depth estimation from 2D images using GIST and saliency. In: "IEEE International Conference on Image Processing (ICIP 2015)", 27/09/2015 - 30/09/2015, Quebec City, Canada. pp. 4753-4757.

Description

Title: Learning-based depth estimation from 2D images using GIST and saliency
Author/s:
  • Arciniegas Herrera, Jose Luis
  • Konrad, Janusz
  • Blanco Adán, Carlos Roberto del
  • García Santos, Narciso
Item Type: Presentation at Congress or Conference (Unspecified)
Event Title: IEEE International Conference on Image Processing (ICIP 2015)
Event Dates: 27/09/2015 - 30/09/2015
Event Location: Quebec City, Canada
Title of Book: IEEE International Conference on Image Processing (ICIP 2015)
Título de Revista/Publicación: 2015 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP)
Date: 2015
ISSN: 1522-4880
Subjects:
Freetext Keywords: 2D-to-3D Image Conversion, Depth maps, GIST Descriptor, Saliency
Faculty: E.T.S.I. Telecomunicación (UPM)
Department: Ingeniería de Sistemas Telemáticos [hasta 2014]
Creative Commons Licenses: Recognition - No derivative works - Non commercial

Full text

[img]
Preview
PDF - Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (1MB) | Preview

Abstract

Although there has been a significant proliferation of 3D displays in the last decade, the availability of 3D content is still scant compared to the volume of 2D data. To fill this gap, automatic 2D to 3D conversion algorithms are needed. In this paper, we present an automatic approach, inspired by machine learning principles, for estimating the depth of a 2D image. The depth of a query image is inferred from a dataset of color and depth images by searching this repository for images that are photometrically similar to the query. We measure the photometric similarity between two images by comparing their GIST descriptors. Since not all regions in the query image require the same visual attention, we give more weight in the GIST-descriptor comparison to regions with high saliency. Subsequently, we fuse the depths of the most similar images and adaptively filter the result to obtain a depth estimate. Our experimental results indicate that the proposed algorithm outperforms other state-of-the-art approaches on the commonly-used Kinect-NYU dataset.

Funding Projects

TypeCodeAcronymLeaderTitle
Government of SpainTEC2010-20412UnspecifiedUnspecifiedAñadiendo percepción de profundidad a las comunicaciones visuales (enhanced 3dtv)
Government of SpainTEC2013-48453UnspecifiedUnspecifiedMixed reality over ultra high definition television

More information

Item ID: 41374
DC Identifier: http://oa.upm.es/41374/
OAI Identifier: oai:oa.upm.es:41374
Official URL: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=7351709
Deposited by: Memoria Investigacion
Deposited on: 06 Jul 2016 16:57
Last Modified: 06 Jul 2016 16:57
  • Logo InvestigaM (UPM)
  • Logo GEOUP4
  • Logo Open Access
  • Open Access
  • Logo Sherpa/Romeo
    Check whether the anglo-saxon journal in which you have published an article allows you to also publish it under open access.
  • Logo Dulcinea
    Check whether the spanish journal in which you have published an article allows you to also publish it under open access.
  • Logo de Recolecta
  • Logo del Observatorio I+D+i UPM
  • Logo de OpenCourseWare UPM