Projection-Based Point Convolution for Efficient Point Cloud Segmentation
Understanding point cloud has recently gained huge interests following the development of 3D scanning devices and the accumulation of large-scale 3D data. Most point cloud processing algorithms can be classified as either point-based or voxel-based methods, both of which have severe limitations in p...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9687584/ |
| _version_ | 1798040425012920320 |
|---|---|
| author | Pyunghwan Ahn Juyoung Yang Eojindl Yi Chanho Lee Junmo Kim |
| author_facet | Pyunghwan Ahn Juyoung Yang Eojindl Yi Chanho Lee Junmo Kim |
| author_sort | Pyunghwan Ahn |
| collection | DOAJ |
| description | Understanding point cloud has recently gained huge interests following the development of 3D scanning devices and the accumulation of large-scale 3D data. Most point cloud processing algorithms can be classified as either point-based or voxel-based methods, both of which have severe limitations in processing time or memory, or both. To overcome these limitations, we propose Projection-based Point Convolution (PPConv), a point convolutional module that uses 2D convolutions and multi-layer perceptrons (MLPs) as its components. In PPConv, point features are processed through two branches: point branch and projection branch. Point branch consists of MLPs, while projection branch transforms point features into a 2D feature map and then apply 2D convolutions. As PPConv does not use point-based or voxel-based convolutions, it has advantages in fast point cloud processing. When combined with a learnable projection and effective feature fusion strategy, PPConv achieves superior efficiency compared to state-of-the-art methods, even with a simple architecture based on PointNet++. We demonstrate the efficiency of PPConv in terms of the trade-off between inference time and segmentation performance. The experimental results on S3DIS and ShapeNetPart show that PPConv is the most efficient method among the compared ones. The code is available at github.com/pahn04/PPConv. |
| first_indexed | 2024-04-11T22:07:26Z |
| format | Article |
| id | doaj.art-c5c854fb7a1945ca960ef8132c2ec521 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-11T22:07:26Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-c5c854fb7a1945ca960ef8132c2ec5212022-12-22T04:00:40ZengIEEEIEEE Access2169-35362022-01-0110153481535810.1109/ACCESS.2022.31444499687584Projection-Based Point Convolution for Efficient Point Cloud SegmentationPyunghwan Ahn0https://orcid.org/0000-0001-6532-5963Juyoung Yang1https://orcid.org/0000-0001-9640-2116Eojindl Yi2https://orcid.org/0000-0002-8036-0641Chanho Lee3Junmo Kim4https://orcid.org/0000-0002-7174-7932School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South KoreaSchool of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South KoreaSchool of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South KoreaSchool of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South KoreaSchool of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South KoreaUnderstanding point cloud has recently gained huge interests following the development of 3D scanning devices and the accumulation of large-scale 3D data. Most point cloud processing algorithms can be classified as either point-based or voxel-based methods, both of which have severe limitations in processing time or memory, or both. To overcome these limitations, we propose Projection-based Point Convolution (PPConv), a point convolutional module that uses 2D convolutions and multi-layer perceptrons (MLPs) as its components. In PPConv, point features are processed through two branches: point branch and projection branch. Point branch consists of MLPs, while projection branch transforms point features into a 2D feature map and then apply 2D convolutions. As PPConv does not use point-based or voxel-based convolutions, it has advantages in fast point cloud processing. When combined with a learnable projection and effective feature fusion strategy, PPConv achieves superior efficiency compared to state-of-the-art methods, even with a simple architecture based on PointNet++. We demonstrate the efficiency of PPConv in terms of the trade-off between inference time and segmentation performance. The experimental results on S3DIS and ShapeNetPart show that PPConv is the most efficient method among the compared ones. The code is available at github.com/pahn04/PPConv.https://ieeexplore.ieee.org/document/9687584/Computer visionobject segmentationrobot vision systems |
| spellingShingle | Pyunghwan Ahn Juyoung Yang Eojindl Yi Chanho Lee Junmo Kim Projection-Based Point Convolution for Efficient Point Cloud Segmentation IEEE Access Computer vision object segmentation robot vision systems |
| title | Projection-Based Point Convolution for Efficient Point Cloud Segmentation |
| title_full | Projection-Based Point Convolution for Efficient Point Cloud Segmentation |
| title_fullStr | Projection-Based Point Convolution for Efficient Point Cloud Segmentation |
| title_full_unstemmed | Projection-Based Point Convolution for Efficient Point Cloud Segmentation |
| title_short | Projection-Based Point Convolution for Efficient Point Cloud Segmentation |
| title_sort | projection based point convolution for efficient point cloud segmentation |
| topic | Computer vision object segmentation robot vision systems |
| url | https://ieeexplore.ieee.org/document/9687584/ |
| work_keys_str_mv | AT pyunghwanahn projectionbasedpointconvolutionforefficientpointcloudsegmentation AT juyoungyang projectionbasedpointconvolutionforefficientpointcloudsegmentation AT eojindlyi projectionbasedpointconvolutionforefficientpointcloudsegmentation AT chanholee projectionbasedpointconvolutionforefficientpointcloudsegmentation AT junmokim projectionbasedpointconvolutionforefficientpointcloudsegmentation |