Sorry, you need to enable JavaScript to visit this website.

SINGLE DEPTH IMAGE SUPER-RESOLUTION USING CONVOLUTIONAL NEURAL NETWORKS

Citation Author(s):
Cheolkon Jung
Submitted by:
Baoliang Chen
Last updated:
12 April 2018 - 11:53pm
Document Type:
Poster
Document Year:
2018
Event:
Presenters:
Baoliang Chen
Paper Code:
ICASSP18001
 

In this paper, a novel framework for the single depth image superresolution is proposed. In our framework, we first extract a low-quality edge map from an interpolated depth map.Then we transform the low-quality edge map to a high quality one by our trained deep convolution neural network (CNN) with two-step postprocessing. Guided by the high-quality edge map, we finally utilize a total variation (TV) based model to upsample the initial depth map. The high quality edge-based guidance not only helps avoiding artifacts introduced by direct texture prediction, but also reduces jagged artifacts and preserves the sharp edges. Experimental results demonstrate the effectiveness of our method both qualitatively and quantitatively compared with the state-of-the-art methods.

up
0 users have voted: