BEVDistill: Cross-Modal BEV Distillation for Multi-View 3D Object Detection

TL;DR

BEVDistill introduces a cross-modal knowledge distillation framework that leverages LiDAR data to enhance multi-view 3D object detection from images, achieving state-of-the-art results without extra inference costs.

Contribution

The paper proposes BEVDistill, a novel cross-modal BEV knowledge distillation method that improves image-based 3D detection by incorporating LiDAR features in a teacher-student paradigm.

Findings

Outperforms existing KD methods on BEVFormer baseline

Achieves 59.4 NDS on nuScenes test leaderboard

No additional inference cost introduced

Abstract

3D object detection from multiple image views is a fundamental and challenging task for visual scene understanding. Owing to its low cost and high efficiency, multi-view 3D object detection has demonstrated promising application prospects. However, accurately detecting objects through perspective views is extremely difficult due to the lack of depth information. Current approaches tend to adopt heavy backbones for image encoders, making them inapplicable for real-world deployment. Different from the images, LiDAR points are superior in providing spatial cues, resulting in highly precise localization. In this paper, we explore the incorporation of LiDAR-based detectors for multi-view 3D object detection. Instead of directly training a depth prediction network, we unify the image and LiDAR features in the Bird-Eye-View (BEV) space and adaptively transfer knowledge across non-homogenous representations in a teacher-student paradigm. To this end, we propose \textbf{BEVDistill}, a cross-modal BEV knowledge distillation (KD) framework for multi-view 3D object detection. Extensive experiments demonstrate that the proposed method outperforms current KD approaches on a highly-competitive baseline, BEVFormer, without introducing any extra cost in the inference phase. Notably, our best model achieves 59.4 NDS on the nuScenes test leaderboard, achieving new state-of-the-art in comparison with various image-based detectors. Code will be available at https://github.com/zehuichen123/BEVDistill.