A Resource Efficient Fusion Network for Object Detection in Bird's-Eye View using Camera and Raw Radar Data

TL;DR

This paper proposes a resource-efficient fusion network that combines camera BEV features and raw radar RD spectrum data for improved object detection in autonomous driving, emphasizing low computational cost.

Contribution

It introduces a novel fusion approach that directly utilizes raw radar data and BEV camera features, reducing processing complexity while maintaining detection accuracy.

Findings

Fusion of BEV camera features with raw radar data improves detection performance.

The proposed method achieves competitive accuracy with lower computational complexity.

Evaluation on RADIal dataset demonstrates effectiveness over existing methods.

Abstract

Cameras can be used to perceive the environment around the vehicle, while affordable radar sensors are popular in autonomous driving systems as they can withstand adverse weather conditions unlike cameras. However, radar point clouds are sparser with low azimuth and elevation resolution that lack semantic and structural information of the scenes, resulting in generally lower radar detection performance. In this work, we directly use the raw range-Doppler (RD) spectrum of radar data, thus avoiding radar signal processing. We independently process camera images within the proposed comprehensive image processing pipeline. Specifically, first, we transform the camera images to Bird's-Eye View (BEV) Polar domain and extract the corresponding features with our camera encoder-decoder architecture. The resultant feature maps are fused with Range-Azimuth (RA) features, recovered from the RD spectrum input from the radar decoder to perform object detection. We evaluate our fusion strategy with other existing methods not only in terms of accuracy but also on computational complexity metrics on RADIal dataset.