Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RADE-Net: Robust Attention Network for Radar-Only Object Detection in Adverse Weather

Christof Leitgeb, Thomas Puchleitner, Max Peter Ronecker, Daniel Watzenig

TL;DR

RADE-Net is introduced, a lightweight model tailored to 3D projections of the RADE tensor, which outperform several Lidar approaches in scenarios with adverse weather conditions and achieves a 16.7% improvement compared to their baseline, as well as 6.5% improvement over current Radar-only models.

Abstract

Automotive perception systems are obligated to meet high requirements. While optical sensors such as Camera and Lidar struggle in adverse weather conditions, Radar provides a more robust perception performance, effectively penetrating fog, rain, and snow. Since full Radar tensors have large data sizes and very few datasets provide them, most Radar-based approaches work with sparse point clouds or 2D projections, which can result in information loss. Additionally, deep learning methods show potential to extract richer and more dense features from low level Radar data and therefore significantly increase the perception performance. Therefore, we propose a 3D projection method for fast-Fourier-transformed 4D Range-Azimuth-Doppler-Elevation (RADE) tensors. Our method preserves rich Doppler and Elevation features while reducing the required data size for a single frame by 91.9% compared to a full tensor, thus achieving higher training and inference speed as well as lower model complexity. We introduce RADE-Net, a lightweight model tailored to 3D projections of the RADE tensor. The backbone enables exploitation of low-level and high-level cues of Radar tensors with spatial and channel-attention. The decoupled detection heads predict object center-points directly in the Range-Azimuth domain and regress rotated 3D bounding boxes from rich feature maps in the cartesian scene. We evaluate the model on scenes with multiple different road users and under various weather conditions on the large-scale K-Radar dataset and achieve a 16.7% improvement compared to their baseline, as well as 6.5% improvement over current Radar-only models. Additionally, we outperform several Lidar approaches in scenarios with adverse weather conditions. The code is available under https://github.com/chr-is-tof/RADE-Net.

RADE-Net: Robust Attention Network for Radar-Only Object Detection in Adverse Weather

TL;DR

RADE-Net is introduced, a lightweight model tailored to 3D projections of the RADE tensor, which outperform several Lidar approaches in scenarios with adverse weather conditions and achieves a 16.7% improvement compared to their baseline, as well as 6.5% improvement over current Radar-only models.

Abstract

Automotive perception systems are obligated to meet high requirements. While optical sensors such as Camera and Lidar struggle in adverse weather conditions, Radar provides a more robust perception performance, effectively penetrating fog, rain, and snow. Since full Radar tensors have large data sizes and very few datasets provide them, most Radar-based approaches work with sparse point clouds or 2D projections, which can result in information loss. Additionally, deep learning methods show potential to extract richer and more dense features from low level Radar data and therefore significantly increase the perception performance. Therefore, we propose a 3D projection method for fast-Fourier-transformed 4D Range-Azimuth-Doppler-Elevation (RADE) tensors. Our method preserves rich Doppler and Elevation features while reducing the required data size for a single frame by 91.9% compared to a full tensor, thus achieving higher training and inference speed as well as lower model complexity. We introduce RADE-Net, a lightweight model tailored to 3D projections of the RADE tensor. The backbone enables exploitation of low-level and high-level cues of Radar tensors with spatial and channel-attention. The decoupled detection heads predict object center-points directly in the Range-Azimuth domain and regress rotated 3D bounding boxes from rich feature maps in the cartesian scene. We evaluate the model on scenes with multiple different road users and under various weather conditions on the large-scale K-Radar dataset and achieve a 16.7% improvement compared to their baseline, as well as 6.5% improvement over current Radar-only models. Additionally, we outperform several Lidar approaches in scenarios with adverse weather conditions. The code is available under https://github.com/chr-is-tof/RADE-Net.
Paper Structure (25 sections, 17 equations, 4 figures, 3 tables)

This paper contains 25 sections, 17 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Background and Related Works
  3. Radar Signal Processing
  4. Radar Object Detection
  5. Model Prediction
  6. Proposed Methodology
  7. Overall Architecture
  8. Tensor Projection Module
  9. Backbone
  10. Dilated Residual Neck
  11. Detection Heads
  12. Loss Function
  13. Focal Loss
  14. Gaussian-Wasserstein Distance Loss
  15. Smooth L1 Loss
  16. ...and 10 more sections

Figures (4)

  • Figure 1: Proposed architecture: (I) Tensor Projection Module (TPM), (II) Encoder-Decoder Backbone, (III) Dilated Neck and (IV) Range-Azimuth based Centerpoint Detection Heads + Cartesian Transformation.
  • Figure 2: Tensor Projection Module (TPM): The full FFT-processed Radar tensor is split into Range-Azimuth-Doppler (RAD) and Range-Azimuth-Elevation (RAE) projections by calculating the maximum along each excluded dimension. The projections are subsequently concatenated along the third dimension.
  • Figure 3: RADE-Net backbone with three down-sampling stages and CBAM supported skip connections.
  • Figure 4: Model performance visualization under different weather conditions. The ground truth bounding boxes are shown in white along with class label and the model predictions are shown in red along with class label and confidence score. The ROI where detections are considered is marked with a dashed black rectangle.