3DSF-MixNet: Mixer-Based Symmetric Scene Flow Estimation From 3D Point Clouds

Wang，Shuaijun1,2; Gao，Rui2; Han，Ruihua2,3; Hao，Qi4

2024

10.1109/LRA.2023.3335776

IEEE Robotics and Automation Letters   影响因子和分区

2377-3774

2377-3766

The scene flow estimation aims at accurately achieving the motion of 3D points, imposing challenges like mis-registration, object occlusions, and non-uniform upsampling. This paper introduces a scene flow estimation framework featuring a unified scene flow estimator, a symmetric cost volume approach, and a geometric/semantic feature based upsampling strategy. The novelty of this work is threefold: (1) developing a novel progressive framework which integrates the cost volume module and scene flow estimator, enhancing scene flow estimation; (2) developing a symmetric inter-frame correlation feature extraction method through cost volume estimation using MLP-Mixer operations; (3) developing an upsampling strategy based on both the semantic and geometric feature similarities between sparse and dense samples. Experimental results show that our method outperforms state-of-the-art baseline methods, especially in scenarios involving challenging conditions, the improvements of our method achieving at most 0.109, m/0.089,m/0.091,m in EPE3D, 54.23% 53.67%/74.1% in AS, 32.75%/21.87%/40.25% in AR, and 70.98%/58.06%/43.56% in outliers, when tested on FlyingThings3D ( FT3D S, FT3D H) and KITTI H datasets, respectively.

Point clouds scene flow

SCI ; EI

英语

通讯

2-s2.0-85178038303

Scopus

1.Harbin Institute of Technology,Department of Computer Science and Engineering,Harbin,150001,China
2.Southern University of Science and Technology,Department of Computer Science and Engineering,Shenzhen,518055,China
3.The University of Hong Kong,Computer Science,Pok Fu Lam,999077,Hong Kong
4.Southern University of Science and Technology,Department of Computer Science and Engineering,Shenzhen Key Laboratory of Robotics and Computer Vision,The Sifakis Research Institute for Trustworthy Autonomous Systems,Shenzhen,518055,China

Wang，Shuaijun,Gao，Rui,Han，Ruihua,et al. 3DSF-MixNet: Mixer-Based Symmetric Scene Flow Estimation From 3D Point Clouds[J]. IEEE Robotics and Automation Letters,2024,9(1):611-618.

Wang，Shuaijun,Gao，Rui,Han，Ruihua,&Hao，Qi.(2024).3DSF-MixNet: Mixer-Based Symmetric Scene Flow Estimation From 3D Point Clouds.IEEE Robotics and Automation Letters,9(1),611-618.

Wang，Shuaijun,et al."3DSF-MixNet: Mixer-Based Symmetric Scene Flow Estimation From 3D Point Clouds".IEEE Robotics and Automation Letters 9.1(2024):611-618.