This paper presents a novel method for view synthesis from sparse neighboring views based on RGB-D images. However, distortion and misalignment can occur in the warped images from neighboring viewpoints to the target viewpoint due to errors in the estimated depth map and camera calibration. This degrades the quality of the novel view images, especially when the cameras are far apart. We propose a novel network to estimate the offsets needed to compensate for the misalignment. The experimental results on the ScanNet dataset show that the proposed method outperforms the others. The experimental results on the ScanNet dataset show that the proposed method outperforms the state-of-the-art methods.
Truong, AM, Philips, W & Deligiannis, N 2022, Novel view synthesis for RGB-D camera networks. in 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP). IVMSP 2022 - 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop, IEEE, pp. 1-5, 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP), Nafplio, Greece, 26/06/22. https://doi.org/10.1109/IVMSP54334.2022.9816256
Truong, A. M., Philips, W., & Deligiannis, N. (2022). Novel view synthesis for RGB-D camera networks. In 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP) (pp. 1-5). (IVMSP 2022 - 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop). IEEE. https://doi.org/10.1109/IVMSP54334.2022.9816256
@inproceedings{0b16966304bd406487058898632eea27,
title = "Novel view synthesis for RGB-D camera networks",
abstract = "This paper presents a novel method for view synthesis from sparse neighboring views based on RGB-D images. However, distortion and misalignment can occur in the warped images from neighboring viewpoints to the target viewpoint due to errors in the estimated depth map and camera calibration. This degrades the quality of the novel view images, especially when the cameras are far apart. We propose a novel network to estimate the offsets needed to compensate for the misalignment. The experimental results on the ScanNet dataset show that the proposed method outperforms the others. The experimental results on the ScanNet dataset show that the proposed method outperforms the state-of-the-art methods.",
author = "Truong, {Anh Minh} and Wilfried Philips and Nikos Deligiannis",
note = "Funding Information: This work was financially supported by the Flemish Fund for Scientific Research FWO-Flanders through the grant 3G014718. Publisher Copyright: {\textcopyright} 2022 IEEE. Copyright: Copyright 2022 Elsevier B.V., All rights reserved.; 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP) ; Conference date: 26-06-2022 Through 29-06-2022",
year = "2022",
doi = "10.1109/IVMSP54334.2022.9816256",
language = "English",
series = "IVMSP 2022 - 2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop",
publisher = "IEEE",
pages = "1--5",
booktitle = "2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP)",
url = "https://2022.ivmsp.org",
}