Multi-view 3D Object Detection Network for Autonomous Driving

TLDR

The paper targets high‑accuracy 3D object detection for autonomous driving. It introduces MV3D, a sensory‑fusion network that encodes sparse LIDAR point clouds into a compact multi‑view representation, generates 3D proposals from a bird‑eye view, and fuses RGB and point‑cloud features across two subnetworks to predict oriented 3D bounding boxes. On the KITTI benchmark, MV3D surpasses state‑of‑the‑art by about 25–30% in 3D localization and detection AP, and by 14.9% in hard‑data 2D detection AP among LIDAR‑based methods.

Abstract

This paper aims at high-accuracy 3D object detection in autonomous driving scenario. We propose Multi-View 3D networks (MV3D), a sensory-fusion framework that takes both LIDAR point cloud and RGB images as input and predicts oriented 3D bounding boxes. We encode the sparse 3D point cloud with a compact multi-view representation. The network is composed of two subnetworks: one for 3D object proposal generation and another for multi-view feature fusion. The proposal network generates 3D candidate boxes efficiently from the birds eye view representation of 3D point cloud. We design a deep fusion scheme to combine region-wise features from multiple views and enable interactions between intermediate layers of different paths. Experiments on the challenging KITTI benchmark show that our approach outperforms the state-of-the-art by around 25% and 30% AP on the tasks of 3D localization and 3D detection. In addition, for 2D detection, our approach obtains 14.9% higher AP than the state-of-the-art on the hard data among the LIDAR-based methods.