ESGN: Efficient Stereo Geometry Network for Fast 3D Object Detection

TL;DR

This paper introduces ESGN, a fast and efficient stereo geometry network that improves 3D object detection accuracy by generating geometry-aware features without complex aggregation, outperforming previous methods on KITTI.

Contribution

The paper proposes a novel efficient stereo geometry network with a geometry-aware feature generation module and a distillation scheme, enhancing accuracy without sacrificing speed.

Findings

ESGN outperforms YOLOStereo3D by 5.14% mAP on KITTI.

ESGN achieves a good balance between detection accuracy and inference speed.

The method effectively utilizes multi-scale stereo volumes and deep feature fusion.

Abstract

Fast stereo based 3D object detectors have made great progress recently. However, they lag far behind high-precision stereo based methods in accuracy. We argue that the main reason is due to the poor geometry-aware feature representation in 3D space. To solve this problem, we propose an efficient stereo geometry network (ESGN). The key in our ESGN is an efficient geometry-aware feature generation (EGFG) module. Our EGFG module first uses a stereo correlation and reprojection module to construct multi-scale stereo volumes in camera frustum space, second employs a multi-scale BEV projection and fusion module to generate multiple geometry-aware features. In these two steps, we adopt deep multi-scale information fusion for discriminative geometry-aware feature generation, without any complex aggregation networks. In addition, we introduce a deep geometry-aware feature distillation scheme to guide stereo feature learning with a LiDAR-based detector. The experiments are performed on the classical KITTI dataset. On KITTI test set, our ESGN outperforms the fast state-of-art-art detector YOLOStereo3D by 5.14\% on mAP$_{3d}$ at 62$ms$. To the best of our knowledge, our ESGN achieves a best trade-off between accuracy and speed. We hope that our efficient stereo geometry network can provide more possible directions for fast 3D object detection. Our source code will be released.