Home > Research > Publications & Outputs > Bridging the Domain Gap in Scene Flow Estimatio...

Links

Text available via DOI:

View graph of relations

Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement. / Zhang, Dejun; Zhang, Mian; Tan, Xuefeng et al.
In: ACM Transactions on Multimedia Computing, Communications, and Applications, Vol. 20, No. 8, 236, 31.08.2024, p. 1-21.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Zhang, D, Zhang, M, Tan, X & Liu, J 2024, 'Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement', ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 20, no. 8, 236, pp. 1-21. https://doi.org/10.1145/3661823

APA

Zhang, D., Zhang, M., Tan, X., & Liu, J. (2024). Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement. ACM Transactions on Multimedia Computing, Communications, and Applications, 20(8), 1-21. Article 236. https://doi.org/10.1145/3661823

Vancouver

Zhang D, Zhang M, Tan X, Liu J. Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement. ACM Transactions on Multimedia Computing, Communications, and Applications. 2024 Aug 31;20(8):1-21. 236. Epub 2024 Jun 12. doi: 10.1145/3661823

Author

Zhang, Dejun ; Zhang, Mian ; Tan, Xuefeng et al. / Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement. In: ACM Transactions on Multimedia Computing, Communications, and Applications. 2024 ; Vol. 20, No. 8. pp. 1-21.

Bibtex

@article{0c4baa3d0ec94b0bb1d32458b19468d7,
title = "Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement",
abstract = "This article introduces SmoothFlowNet3D, an innovative encoder-decoder architecture specifically designed for bridging the domain gap in scene flow estimation. To achieve this goal, SmoothFlowNet3D divides the scene flow estimation task into two stages: initial scene flow estimation and smoothness refinement. Specifically, SmoothFlowNet3D comprises a hierarchical encoder that extracts multi-scale point cloud features from two consecutive frames, along with a hierarchical decoder responsible for predicting the initial scene flow and further refining it to achieve smoother estimation. To generate the initial scene flow, a cross-frame nearest-neighbor search operation is performed between the features extracted from two consecutive frames, resulting in forward and backward flow embeddings. These embeddings are then combined to form the bidirectional flow embedding, serving as input for predicting the initial scene flow. Additionally, a flow smoothing module based on the self-attention mechanism is proposed to predict the smoothing error and facilitate the refinement of the initial scene flow for more accurate and smoother estimation results. Extensive experiments demonstrate that the proposed SmoothFlowNet3D approach achieves state-of-the-art performance on both synthetic datasets and real LiDAR point clouds, confirming its effectiveness in enhancing scene flow smoothness.",
author = "Dejun Zhang and Mian Zhang and Xuefeng Tan and Jun Liu",
year = "2024",
month = aug,
day = "31",
doi = "10.1145/3661823",
language = "English",
volume = "20",
pages = "1--21",
journal = "ACM Transactions on Multimedia Computing, Communications, and Applications",
issn = "1551-6857",
publisher = "Association for Computing Machinery (ACM)",
number = "8",

}

RIS

TY - JOUR

T1 - Bridging the Domain Gap in Scene Flow Estimation via Hierarchical Smoothness Refinement

AU - Zhang, Dejun

AU - Zhang, Mian

AU - Tan, Xuefeng

AU - Liu, Jun

PY - 2024/8/31

Y1 - 2024/8/31

N2 - This article introduces SmoothFlowNet3D, an innovative encoder-decoder architecture specifically designed for bridging the domain gap in scene flow estimation. To achieve this goal, SmoothFlowNet3D divides the scene flow estimation task into two stages: initial scene flow estimation and smoothness refinement. Specifically, SmoothFlowNet3D comprises a hierarchical encoder that extracts multi-scale point cloud features from two consecutive frames, along with a hierarchical decoder responsible for predicting the initial scene flow and further refining it to achieve smoother estimation. To generate the initial scene flow, a cross-frame nearest-neighbor search operation is performed between the features extracted from two consecutive frames, resulting in forward and backward flow embeddings. These embeddings are then combined to form the bidirectional flow embedding, serving as input for predicting the initial scene flow. Additionally, a flow smoothing module based on the self-attention mechanism is proposed to predict the smoothing error and facilitate the refinement of the initial scene flow for more accurate and smoother estimation results. Extensive experiments demonstrate that the proposed SmoothFlowNet3D approach achieves state-of-the-art performance on both synthetic datasets and real LiDAR point clouds, confirming its effectiveness in enhancing scene flow smoothness.

AB - This article introduces SmoothFlowNet3D, an innovative encoder-decoder architecture specifically designed for bridging the domain gap in scene flow estimation. To achieve this goal, SmoothFlowNet3D divides the scene flow estimation task into two stages: initial scene flow estimation and smoothness refinement. Specifically, SmoothFlowNet3D comprises a hierarchical encoder that extracts multi-scale point cloud features from two consecutive frames, along with a hierarchical decoder responsible for predicting the initial scene flow and further refining it to achieve smoother estimation. To generate the initial scene flow, a cross-frame nearest-neighbor search operation is performed between the features extracted from two consecutive frames, resulting in forward and backward flow embeddings. These embeddings are then combined to form the bidirectional flow embedding, serving as input for predicting the initial scene flow. Additionally, a flow smoothing module based on the self-attention mechanism is proposed to predict the smoothing error and facilitate the refinement of the initial scene flow for more accurate and smoother estimation results. Extensive experiments demonstrate that the proposed SmoothFlowNet3D approach achieves state-of-the-art performance on both synthetic datasets and real LiDAR point clouds, confirming its effectiveness in enhancing scene flow smoothness.

U2 - 10.1145/3661823

DO - 10.1145/3661823

M3 - Journal article

VL - 20

SP - 1

EP - 21

JO - ACM Transactions on Multimedia Computing, Communications, and Applications

JF - ACM Transactions on Multimedia Computing, Communications, and Applications

SN - 1551-6857

IS - 8

M1 - 236

ER -