DANet: Density Adaptive Convolutional Network with Interactive Attention for 3D Point Clouds
This work improves 3D point cloud processing for applications like computer vision and robotics by introducing a novel density-adaptive convolution and efficient attention module, though it is incremental in advancing existing methods.
The paper tackles the problem of 3D point cloud analysis by addressing robustness to varying point density and computational inefficiency in contextual modeling, achieving state-of-the-art classification results of 93.6% on ModelNet40 and competitive segmentation results.
Local features and contextual dependencies are crucial for 3D point cloud analysis. Many works have been devoted to designing better local convolutional kernels that exploit the contextual dependencies. However, current point convolutions lack robustness to varying point cloud density. Moreover, contextual modeling is dominated by non-local or self-attention models which are computationally expensive. To solve these problems, we propose density adaptive convolution, coined DAConv. The key idea is to adaptively learn the convolutional weights from geometric connections obtained from the point density and position. To extract precise context dependencies with fewer computations, we propose an interactive attention module (IAM) that embeds spatial information into channel attention along different spatial directions. DAConv and IAM are integrated in a hierarchical network architecture to achieve local density and contextual direction-aware learning for point cloud analysis. Experiments show that DAConv is significantly more robust to point density compared to existing methods and extensive comparisons on challenging 3D point cloud datasets show that our network achieves state-of-the-art classification results of 93.6% on ModelNet40, competitive semantic segmentation results of 68.71% mIoU on S3DIS and part segmentation results of 86.7% mIoU on ShapeNet.