SpikingRTNH: Spiking Neural Network for 4D Radar Object Detection

2025-01-31

AI Summary

To address the high computational energy consumption and insufficient robustness under adverse weather conditions of dense 4D radar point clouds in autonomous driving 3D object detection, this paper proposes the first spiking neural network (SNN)-based lightweight and efficient detection framework. We introduce a novel SNN architecture specifically designed for 4D radar data, incorporating a biologically inspired top-down inference (BTI) mechanism that sequentially processes points according to their density-based priority—thereby significantly enhancing signal-to-noise ratio and key-point utilization. The framework integrates leaky integrate-and-fire (LIF) neurons, customized point cloud encoding, and an optimized detection head. Evaluated on the K-Radar dataset, our method achieves 51.1% 3D average precision (AP) and 57.0% bird’s-eye view (BEV) AP—matching state-of-the-art artificial neural network (ANN) baselines—while reducing power consumption by 78%. The approach thus delivers low-power operation, real-time inference capability, and all-weather robustness.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Computer Vision: Object Detection & Categorization

Application Category

• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Recently, 4D Radar has emerged as a crucial sensor for 3D object detection in autonomous vehicles, offering both stable perception in adverse weather and high-density point clouds for object shape recognition. However, processing such high-density data demands substantial computational resources and energy consumption. We propose SpikingRTNH, the first spiking neural network (SNN) for 3D object detection using 4D Radar data. By replacing conventional ReLU activation functions with leaky integrate-and-fire (LIF) spiking neurons, SpikingRTNH achieves significant energy efficiency gains. Furthermore, inspired by human cognitive processes, we introduce biological top-down inference (BTI), which processes point clouds sequentially from higher to lower densities. This approach effectively utilizes points with lower noise and higher importance for detection. Experiments on K-Radar dataset demonstrate that SpikingRTNH with BTI significantly reduces energy consumption by 78% while achieving comparable detection performance to its ANN counterpart (51.1% AP 3D, 57.0% AP BEV). These results establish the viability of SNNs for energy-efficient 4D Radar-based object detection in autonomous driving systems. All codes are available at https://github.com/kaist-avelab/k-radar.

Problem

• Autonomous Vehicles
• 4D Radar Processing
• Resource Efficiency

Innovation

• Spiking Neural Networks
• 4D Radar Object Detection
• Leaky Integrate-and-Fire (LIF) Neurons