Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

SGNet: Structure Guided Network via Gradient-Frequency Awareness for Depth Map Super-Resolution

About

Depth super-resolution (DSR) aims to restore high-resolution (HR) depth from low-resolution (LR) one, where RGB image is often used to promote this task. Recent image guided DSR approaches mainly focus on spatial domain to rebuild depth structure. However, since the structure of LR depth is usually blurry, only considering spatial domain is not very sufficient to acquire satisfactory results. In this paper, we propose structure guided network (SGNet), a method that pays more attention to gradient and frequency domains, both of which have the inherent ability to capture high-frequency structure. Specifically, we first introduce the gradient calibration module (GCM), which employs the accurate gradient prior of RGB to sharpen the LR depth structure. Then we present the Frequency Awareness Module (FAM) that recursively conducts multiple spectrum differencing blocks (SDB), each of which propagates the precise high-frequency components of RGB into the LR depth. Extensive experimental results on both real and synthetic datasets demonstrate the superiority of our SGNet, reaching the state-of-the-art. Codes and pre-trained models are available at https://github.com/yanzq95/SGNet.

Zhengxue Wang, Zhiqiang Yan, Jian Yang• 2023

Related benchmarks

TaskDatasetResultRank
Depth Super-ResolutionTOFDSR
RMSE0.0726
30
Depth Super-ResolutionRGB-D-D
RMSE0.0723
30
RGB-guided depth Super-ResolutionReal-world RGB-D-D x2 (test)
RMSE2.98
17
RGB-guided depth Super-ResolutionMiddlebury x4 scale 2006
RMSE2.33
17
RGB-guided depth Super-ResolutionMiddlebury x8 scale 2006
RMSE3.34
17
Compressed Depth Map Super-ResolutionSynthesized Compressed-NYU (test)
MAE0.0426
10
Compressed Depth UpsamplingAIM Compressed Depth Upsampling Challenge 2024
MAE1.337
9
Showing 7 of 7 rows

Other info

Follow for update