DM3D: Distortion-Minimized Weight Pruning for Lossless 3D Object Detection

TL;DR

This paper introduces DM3D, a novel weight pruning method for 3D object detection models that minimizes detection distortion, enabling significant computational reduction while maintaining or improving detection accuracy.

Contribution

It proposes a universal, post-training pruning framework based on second-order Taylor approximation that minimizes detection distortion in 3D object detection models.

Findings

Achieves over 3.89x FLOPs reduction on CenterPoint without accuracy loss.

Maintains or improves detection precision after pruning.

Outperforms state-of-the-art pruning methods in 3D detection tasks.

Abstract

Applying deep neural networks to 3D point cloud processing has attracted increasing attention due to its advanced performance in many areas, such as AR/VR, autonomous driving, and robotics. However, as neural network models and 3D point clouds expand in size, it becomes a crucial challenge to reduce the computational and memory overhead to meet latency and energy constraints in real-world applications. Although existing approaches have proposed to reduce both computational cost and memory footprint, most of them only address the spatial redundancy in inputs, i.e. removing the redundancy of background points in 3D data. In this paper, we propose a novel post-training weight pruning scheme for 3D object detection that is (1) orthogonal to all existing point cloud sparsifying methods, which determines redundant parameters in the pretrained model that lead to minimal distortion in both locality and confidence (detection distortion); and (2) a universal plug-and-play pruning framework that works with arbitrary 3D detection model. This framework aims to minimize detection distortion of network output to maximally maintain detection precision, by identifying layer-wise sparsity based on second-order Taylor approximation of the distortion. Albeit utilizing second-order information, we introduced a lightweight scheme to efficiently acquire Hessian information, and subsequently perform dynamic programming to solve the layer-wise sparsity. Extensive experiments on KITTI, Nuscenes and ONCE datasets demonstrate that our approach is able to maintain and even boost the detection precision on pruned model under noticeable computation reduction (FLOPs). Noticeably, we achieve over 3.89x, 3.72x FLOPs reduction on CenterPoint and PVRCNN model, respectively, without mAP decrease, significantly improving the state-of-the-art.