To the Point: Efficient 3D Object Detection in the Range Image with Graph Convolution Kernels

TL;DR

This paper introduces a novel 3D object detection method using range images and graph convolution kernels, achieving high accuracy with significantly reduced computational requirements.

Contribution

It proposes a new 2D convolutional network architecture with flexible kernels for 3D detection from range images, outperforming existing methods in efficiency and accuracy.

Findings

Outperforms state-of-the-art on Waymo dataset

Achieves 75.5% AP for pedestrians

Requires 180 times fewer FLOPS and parameters

Abstract

3D object detection is vital for many robotics applications. For tasks where a 2D perspective range image exists, we propose to learn a 3D representation directly from this range image view. To this end, we designed a 2D convolutional network architecture that carries the 3D spherical coordinates of each pixel throughout the network. Its layers can consume any arbitrary convolution kernel in place of the default inner product kernel and exploit the underlying local geometry around each pixel. We outline four such kernels: a dense kernel according to the bag-of-words paradigm, and three graph kernels inspired by recent graph neural network advances: the Transformer, the PointNet, and the Edge Convolution. We also explore cross-modality fusion with the camera image, facilitated by operating in the perspective range image view. Our method performs competitively on the Waymo Open Dataset and improves the state-of-the-art AP for pedestrian detection from 69.7% to 75.5%. It is also efficient in that our smallest model, which still outperforms the popular PointPillars in quality, requires 180 times fewer FLOPS and model parameters