Points to Patches: Enabling the Use of Self-Attention for 3D Shape Recognition
Authors:
- 21 August 2022
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 528-534
Abstract
While the Transformer architecture has become ubiquitous in the machine learning field, its adaptation to 3D shape recognition is non-trivial. Due to its quadratic computational complexity, the self-attention operator quickly becomes inefficient as the set of input points grows larger. Furthermore, we find that the attention mechanism struggles to find useful connections between individual points on a global scale. In order to alleviate these problems, we propose a two-stage Point Transformer-in-Transformer (Point-TnT) approach which combines local and global attention mechanisms, enabling both individual points and patches of points to attend to each other effectively. Experiments on shape classification show that such an approach provides more useful features for downstream tasks than the baseline Transformer, while also being more computationally efficient. In addition, we also extend our method to feature matching for scene reconstruction, showing that it can be used in conjunction with existing scene reconstruction pipelines.Keywords
All Related Versions
This publication has 20 references indexed in Scilit:
- Conformer: Convolution-augmented Transformer for Speech Recognition2020
- View-GCN: View-Based Graph Convolutional Network for 3D Shape Analysis2020
- Fully Convolutional Geometric Features2019
- Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data2019
- 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks2019
- SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters2018
- 3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions2017
- Deep Residual Learning for Image Recognition2016
- VoxNet: A 3D Convolutional Neural Network for real-time object recognition2015
- The farthest point strategy for progressive image samplingIEEE Transactions on Image Processing, 1997