Local Grid Rendering Networks for 3D Object Detection in Point Clouds
This work addresses a key bottleneck in 3D object detection for applications like robotics and autonomous driving by enhancing pattern learning in point-based models with minimal computational overhead.
The paper tackles the problem of 3D object detection in point clouds by improving local geometric pattern modeling, proposing a Local Grid Rendering operation and LGR-Net backbone that achieve state-of-the-art results with 5.5 and 4.5 mAP gains on ScanNet and SUN RGB-D datasets, respectively, while maintaining computational efficiency.
The performance of 3D object detection models over point clouds highly depends on their capability of modeling local geometric patterns. Conventional point-based models exploit local patterns through a symmetric function (e.g. max pooling) or based on graphs, which easily leads to loss of fine-grained geometric structures. Regarding capturing spatial patterns, CNNs are powerful but it would be computationally costly to directly apply convolutions on point data after voxelizing the entire point clouds to a dense regular 3D grid. In this work, we aim to improve performance of point-based models by enhancing their pattern learning ability through leveraging CNNs while preserving computational efficiency. We propose a novel and principled Local Grid Rendering (LGR) operation to render the small neighborhood of a subset of input points into a low-resolution 3D grid independently, which allows small-size CNNs to accurately model local patterns and avoids convolutions over a dense grid to save computation cost. With the LGR operation, we introduce a new generic backbone called LGR-Net for point cloud feature extraction with simple design and high efficiency. We validate LGR-Net for 3D object detection on the challenging ScanNet and SUN RGB-D datasets. It advances state-of-the-art results significantly by 5.5 and 4.5 mAP, respectively, with only slight increased computation overhead.