This paper addresses the problem of depth estimation from a single still image. Inspired by recent works on multiscale convolutional neural networks (CNN), we propose a deep model which fuses complementary information derived from multiple CNN side outputs. Different from previous methods, the integration is obtained by means of continuous Conditional Random Fields (CRFs). In particular, we propose two different variations, one based on a cascade of multiple CRFs, the other on a unified graphical model. By designing a novel CNN implementation of mean-field updates for continuous CRFs, we show that both proposed models can be regarded as sequential deep networks and that training can be performed end-to-end. Through extensive experimental evaluation we demonstrate the effectiveness of the proposed approach and establish new state of the art results on publicly available datasets.

Multi-scale Continuous CRFs as Sequential Deep Networks for Monocular Depth Estimation / Xu, Dan; Ricci, Elisa; Ouyang, Wanli; Wang, Xiaogang; Sebe, Nicu. - (2017), pp. 161-169. (Intervento presentato al convegno 30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017) tenutosi a Honolulu HI, USA nel 21-26 JUL, 2017) [10.1109/CVPR.2017.25].

Multi-scale Continuous CRFs as Sequential Deep Networks for Monocular Depth Estimation

Xu, Dan;Ricci, Elisa;Sebe, Nicu
2017-01-01

Abstract

This paper addresses the problem of depth estimation from a single still image. Inspired by recent works on multiscale convolutional neural networks (CNN), we propose a deep model which fuses complementary information derived from multiple CNN side outputs. Different from previous methods, the integration is obtained by means of continuous Conditional Random Fields (CRFs). In particular, we propose two different variations, one based on a cascade of multiple CRFs, the other on a unified graphical model. By designing a novel CNN implementation of mean-field updates for continuous CRFs, we show that both proposed models can be regarded as sequential deep networks and that training can be performed end-to-end. Through extensive experimental evaluation we demonstrate the effectiveness of the proposed approach and establish new state of the art results on publicly available datasets.
2017
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
Piscataway, NJ USA
IEEE
978-1-5386-0457-1
Xu, Dan; Ricci, Elisa; Ouyang, Wanli; Wang, Xiaogang; Sebe, Nicu
Multi-scale Continuous CRFs as Sequential Deep Networks for Monocular Depth Estimation / Xu, Dan; Ricci, Elisa; Ouyang, Wanli; Wang, Xiaogang; Sebe, Nicu. - (2017), pp. 161-169. (Intervento presentato al convegno 30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017) tenutosi a Honolulu HI, USA nel 21-26 JUL, 2017) [10.1109/CVPR.2017.25].
File in questo prodotto:
File Dimensione Formato  
Xu_Multi-Scale_Continuous_CRFs_CVPR_2017_paper.pdf

accesso aperto

Tipologia: Versione editoriale (Publisher’s layout)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 1.15 MB
Formato Adobe PDF
1.15 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11572/193396
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 386
  • ???jsp.display-item.citation.isi??? 288
  • OpenAlex ND
social impact