TY - GEN
T1 - ASSANet
T2 - 35th Conference on Neural Information Processing Systems, NeurIPS 2021
AU - Qian, Guocheng
AU - Al Kader Hammoud, Hasan Abed
AU - Li, Guohao
AU - Thabet, Ali
AU - Ghanem, Bernard
N1 - Funding Information:
The authors appreciate the anonymous NeurIPS reviewers for their constructive feedback (including the revised title, the feature pattern visualization, and the additional experiments). This work was supported by the KAUST Office of Sponsored Research (OSR) through the Visual Computing Center (VCC) funding.
Publisher Copyright:
© 2021 Neural information processing systems foundation. All rights reserved.
PY - 2021
Y1 - 2021
N2 - Access to 3D point cloud representations has been widely facilitated by LiDAR sensors embedded in various mobile devices. This has led to an emerging need for fast and accurate point cloud processing techniques. In this paper, we revisit and dive deeper into PointNet++, one of the most influential yet under-explored networks, and develop faster and more accurate variants of the model. We first present a novel Separable Set Abstraction (SA) module that disentangles the vanilla SA module used in PointNet++ into two separate learning stages: (1) learning channel correlation and (2) learning spatial correlation. The Separable SA module is significantly faster than the vanilla version, yet it achieves comparable performance. We then introduce a new Anisotropic Reduction function into our Separable SA module and propose an Anisotropic Separable SA (ASSA) module that substantially increases the network's accuracy. We later replace the vanilla SA modules in PointNet++ with the proposed ASSA module, and denote the modified network as ASSANet. Extensive experiments on point cloud classification, semantic segmentation, and part segmentation show that ASSANet outperforms PointNet++ and other methods, achieving much higher accuracy and faster speeds. In particular, ASSANet outperforms PointNet++ by 7.4 mIoU on S3DIS Area 5, while maintaining 1.6× faster inference speed on a single NVIDIA 2080Ti GPU. Our scaled ASSANet variant achieves 66.8 mIoU and outperforms KPConv, while being more than 54× faster.
AB - Access to 3D point cloud representations has been widely facilitated by LiDAR sensors embedded in various mobile devices. This has led to an emerging need for fast and accurate point cloud processing techniques. In this paper, we revisit and dive deeper into PointNet++, one of the most influential yet under-explored networks, and develop faster and more accurate variants of the model. We first present a novel Separable Set Abstraction (SA) module that disentangles the vanilla SA module used in PointNet++ into two separate learning stages: (1) learning channel correlation and (2) learning spatial correlation. The Separable SA module is significantly faster than the vanilla version, yet it achieves comparable performance. We then introduce a new Anisotropic Reduction function into our Separable SA module and propose an Anisotropic Separable SA (ASSA) module that substantially increases the network's accuracy. We later replace the vanilla SA modules in PointNet++ with the proposed ASSA module, and denote the modified network as ASSANet. Extensive experiments on point cloud classification, semantic segmentation, and part segmentation show that ASSANet outperforms PointNet++ and other methods, achieving much higher accuracy and faster speeds. In particular, ASSANet outperforms PointNet++ by 7.4 mIoU on S3DIS Area 5, while maintaining 1.6× faster inference speed on a single NVIDIA 2080Ti GPU. Our scaled ASSANet variant achieves 66.8 mIoU and outperforms KPConv, while being more than 54× faster.
UR - http://www.scopus.com/inward/record.url?scp=85128526961&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85128526961
T3 - Advances in Neural Information Processing Systems
SP - 28119
EP - 28130
BT - Advances in Neural Information Processing Systems 34 - 35th Conference on Neural Information Processing Systems, NeurIPS 2021
A2 - Ranzato, Marc'Aurelio
A2 - Beygelzimer, Alina
A2 - Dauphin, Yann
A2 - Liang, Percy S.
A2 - Wortman Vaughan, Jenn
PB - Neural information processing systems foundation
Y2 - 6 December 2021 through 14 December 2021
ER -