Tampere University of Technology

TUTCRIS Research Portal

MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction. / Gao, Yuan; Bregovic, Robert; Gotchev, Atanas; Koch, Reinhard.

2019 IEEE International Conference on Multimedia and Expo, ICME 2019. IEEE, 2019. p. 187-192.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Gao, Y, Bregovic, R, Gotchev, A & Koch, R 2019, MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction. in 2019 IEEE International Conference on Multimedia and Expo, ICME 2019. IEEE, pp. 187-192, IEEE International Conference on Multimedia and Expo, Shanghai, China, 8/07/19. https://doi.org/10.1109/ICME.2019.00040

APA

Gao, Y., Bregovic, R., Gotchev, A., & Koch, R. (2019). MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction. In 2019 IEEE International Conference on Multimedia and Expo, ICME 2019 (pp. 187-192). IEEE. https://doi.org/10.1109/ICME.2019.00040

Vancouver

Gao Y, Bregovic R, Gotchev A, Koch R. MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction. In 2019 IEEE International Conference on Multimedia and Expo, ICME 2019. IEEE. 2019. p. 187-192 https://doi.org/10.1109/ICME.2019.00040

Author

Gao, Yuan ; Bregovic, Robert ; Gotchev, Atanas ; Koch, Reinhard. / MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction. 2019 IEEE International Conference on Multimedia and Expo, ICME 2019. IEEE, 2019. pp. 187-192

Bibtex - Download

@inproceedings{daa371fcbff14332b9ec8881acc4e462,
title = "MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction",
abstract = "Shearlet Transform (ST) is one of the most effective algorithms for the Densely-Sampled Light Field (DSLF) reconstruction from a Sparsely-Sampled Light Field (SSLF) with a large disparity range. However, ST requires a precise estimation of the disparity range of the SSLF in order to design a shearlet system with decent scales and to pre-shear the sparsely-sampled Epipolar-Plane Images (EPIs) of the SSLF. To overcome this limitation, a novel coarse-to-fine DSLF reconstruction method, referred to as Mask-Accelerated Shearlet Transform (MAST), is proposed in this paper. Specifically, a state-of-the-art learning-based optical flow method, FlowNet2, is employed to estimate the disparities of a SSLF. The estimated disparities are then utilized to roughly estimate the densely-sampled EPIs for the sparsely-sampled EPIs of the SSLF. Finally, an elaborately-designed soft mask for a coarsely-inpainted EPI is exploited to perform an iterative refinement on this EPI. Experimental results on nine challenging horizontal-parallax real-world SSLF datasets with large disparity ranges (up to 35 pixels) demonstrate the effectiveness and efficiency of the proposed method over the other state-of-the-art approaches.",
keywords = "Densely-sampled light field reconstruction, Mask-accelerated shearlet transform, Parallax view generation, Shearlet transform, View synthesis",
author = "Yuan Gao and Robert Bregovic and Atanas Gotchev and Reinhard Koch",
year = "2019",
month = "7",
day = "1",
doi = "10.1109/ICME.2019.00040",
language = "English",
pages = "187--192",
booktitle = "2019 IEEE International Conference on Multimedia and Expo, ICME 2019",
publisher = "IEEE",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - MAST: Mask-accelerated shearlet transform for densely-sampled light field reconstruction

AU - Gao, Yuan

AU - Bregovic, Robert

AU - Gotchev, Atanas

AU - Koch, Reinhard

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Shearlet Transform (ST) is one of the most effective algorithms for the Densely-Sampled Light Field (DSLF) reconstruction from a Sparsely-Sampled Light Field (SSLF) with a large disparity range. However, ST requires a precise estimation of the disparity range of the SSLF in order to design a shearlet system with decent scales and to pre-shear the sparsely-sampled Epipolar-Plane Images (EPIs) of the SSLF. To overcome this limitation, a novel coarse-to-fine DSLF reconstruction method, referred to as Mask-Accelerated Shearlet Transform (MAST), is proposed in this paper. Specifically, a state-of-the-art learning-based optical flow method, FlowNet2, is employed to estimate the disparities of a SSLF. The estimated disparities are then utilized to roughly estimate the densely-sampled EPIs for the sparsely-sampled EPIs of the SSLF. Finally, an elaborately-designed soft mask for a coarsely-inpainted EPI is exploited to perform an iterative refinement on this EPI. Experimental results on nine challenging horizontal-parallax real-world SSLF datasets with large disparity ranges (up to 35 pixels) demonstrate the effectiveness and efficiency of the proposed method over the other state-of-the-art approaches.

AB - Shearlet Transform (ST) is one of the most effective algorithms for the Densely-Sampled Light Field (DSLF) reconstruction from a Sparsely-Sampled Light Field (SSLF) with a large disparity range. However, ST requires a precise estimation of the disparity range of the SSLF in order to design a shearlet system with decent scales and to pre-shear the sparsely-sampled Epipolar-Plane Images (EPIs) of the SSLF. To overcome this limitation, a novel coarse-to-fine DSLF reconstruction method, referred to as Mask-Accelerated Shearlet Transform (MAST), is proposed in this paper. Specifically, a state-of-the-art learning-based optical flow method, FlowNet2, is employed to estimate the disparities of a SSLF. The estimated disparities are then utilized to roughly estimate the densely-sampled EPIs for the sparsely-sampled EPIs of the SSLF. Finally, an elaborately-designed soft mask for a coarsely-inpainted EPI is exploited to perform an iterative refinement on this EPI. Experimental results on nine challenging horizontal-parallax real-world SSLF datasets with large disparity ranges (up to 35 pixels) demonstrate the effectiveness and efficiency of the proposed method over the other state-of-the-art approaches.

KW - Densely-sampled light field reconstruction

KW - Mask-accelerated shearlet transform

KW - Parallax view generation

KW - Shearlet transform

KW - View synthesis

U2 - 10.1109/ICME.2019.00040

DO - 10.1109/ICME.2019.00040

M3 - Conference contribution

SP - 187

EP - 192

BT - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019

PB - IEEE

ER -