Transferable Class Statistics and Multi-scale Feature Approximation for 3D Object Detection
This work addresses the problem of high computational resource requirements for lightweight 3D object detection models, which is incremental as it builds on existing methods to improve efficiency.
The paper tackles the computational inefficiency of multi-scale feature learning in 3D object detection from point clouds by approximating multi-scale features from a single neighborhood using knowledge distillation and transferable class statistics, achieving competitive performance on public datasets while reducing computational costs.
This paper investigates multi-scale feature approximation and transferable features for object detection from point clouds. Multi-scale features are critical for object detection from point clouds. However, multi-scale feature learning usually involves multiple neighborhood searches and scale-aware layers, which can hinder efforts to achieve lightweight models and may not be conducive to research constrained by limited computational resources. This paper approximates point-based multi-scale features from a single neighborhood based on knowledge distillation. To compensate for the loss of constructive diversity in a single neighborhood, this paper designs a transferable feature embedding mechanism. Specifically, class-aware statistics are employed as transferable features given the small computational cost. In addition, this paper introduces the central weighted intersection over union for localization to alleviate the misalignment brought by the center offset in optimization. Note that the method presented in this paper saves computational costs. Extensive experiments on public datasets demonstrate the effectiveness of the proposed method.