Skip to main content

Table 2 SSIM of estimated depth on Middlebury datasets when cross-boundary scribbles are present

From: Robust semi-automatic 2D-to-3D image conversion via residual-driven optimization

 

RW

HGR

NRW

OPT

OCP

SDF

â„“ 1

Proposed

Tsukuba

0.724

0.716

0.723

0.727

0.708

0.724

0.722

0.731

Venus

0.969

0.961

0.968

0.970

0.966

0.969

0.971

0.974

Teddy

0.861

0.846

0.865

0.862

0.860

0.861

0.868

0.865

Cones

0.900

0.871

0.900

0.902

0.850

0.897

0.885

0.903

Average

0.864

0.848

0.864

0.865

0.846

0.863

0.862

0.868

  1. The first and second best SSIM at each row are shown in bold and italics, respectively