This paper presents a new approach to boost a single-modality (LiDAR) 3D
object detector by teaching it to simulate features and responses that follow a
multi-modality (LiDAR-image) detector. The approach needs LiDAR-image data only
when training the single-modality detector, and once well-trained, it only
needs LiDAR data at inference. We design a novel framework to realize the
approach: response distillation to focus on the crucial response samples and
avoid the background samples; sparse-voxel distillation to learn voxel
semantics and relations from the estimated crucial voxels; a fine-grained
voxel-to-point distillation to better attend to features of small and distant
objects; and instance distillation to further enhance the deep-feature
consistency. Experimental results on the nuScenes dataset show that our
approach outperforms all SOTA LiDAR-only 3D detectors and even surpasses the
baseline LiDAR-image detector on the key NDS metric, filling 72% mAP gap
between the single- and multi-modality detectors.

该研究提出一种新的方法，通过教会单模（LiDAR）三维目标探测器模拟多模（LiDAR 图像）探测器的特征和响应来提高其效率，从而实现在推断时只需要 LiDAR 数据的目标识别，经过实验，这种方法在性能上优于所有 SOTA 的 LiDAR-only 3D 检测器，甚至超越了基线 LiDAR-image 检测器，填补了单模与多模检测器之间的 72％mAP 差距。