Exploring Data Augmentation for Multi-Modality 3D Object Detection
This work addresses the performance gap in multi-modality 3D object detection for autonomous driving, offering an incremental improvement by enhancing data augmentation techniques.
This paper investigates why multi-modality 3D object detection methods often underperform compared to point cloud-only approaches, attributing it to insufficient data augmentation. They introduce a 'transformation flow' pipeline for consistent multi-modality augmentation and 'MoCa' for occlusion-aware cut-and-paste, achieving new state-of-the-art on nuScenes and competitive results on KITTI.
It is counter-intuitive that multi-modality methods based on point cloud and images perform only marginally better or sometimes worse than approaches that solely use point cloud. This paper investigates the reason behind this phenomenon. Due to the fact that multi-modality data augmentation must maintain consistency between point cloud and images, recent methods in this field typically use relatively insufficient data augmentation. This shortage makes their performance under expectation. Therefore, we contribute a pipeline, named transformation flow, to bridge the gap between single and multi-modality data augmentation with transformation reversing and replaying. In addition, considering occlusions, a point in different modalities may be occupied by different objects, making augmentations such as cut and paste non-trivial for multi-modality detection. We further present Multi-mOdality Cut and pAste (MoCa), which simultaneously considers occlusion and physical plausibility to maintain the multi-modality consistency. Without using ensemble of detectors, our multi-modality detector achieves new state-of-the-art performance on nuScenes dataset and competitive performance on KITTI 3D benchmark. Our method also wins the best PKL award in the 3rd nuScenes detection challenge. Code and models will be released at https://github.com/open-mmlab/mmdetection3d.