Synthetic FMCW Radar Range Azimuth Maps Augmentation with Generative Diffusion Model

2026-01-09

arXiv.org

AI Summary

This study addresses the scarcity and limited diversity of annotated automotive radar data, which hinder the performance of deep learning in environmental perception. To overcome this, the work introduces diffusion models for the first time to generate FMCW radar range-azimuth maps, proposing a conditional generative framework that enables semantic control via confidence maps. By integrating geometry-aware conditioning and temporal consistency regularization, the method ensures physical plausibility and dynamic coherence of the synthesized data. Evaluated on the ROD2021 dataset, the approach improves the peak signal-to-noise ratio of reconstructed signals by 3.6 dB. Furthermore, joint training with synthetic and real data boosts the average precision of downstream detection tasks by 4.15%, significantly enhancing model generalization.

Technology Category

Computer Vision: Diffusion Models for VisionMachine Learning: Deep Generative Models & AutoencodersNatural Language Processing: Generation

Application Category

Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsUser Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

The scarcity and low diversity of well-annotated automotive radar datasets often limit the performance of deep-learning-based environmental perception. To overcome these challenges, we propose a conditional generative framework for synthesizing realistic Frequency-Modulated Continuous-Wave radar Range-Azimuth Maps. Our approach leverages a generative diffusion model to generate radar data for multiple object categories, including pedestrians, cars, and cyclists. Specifically, conditioning is achieved via Confidence Maps, where each channel represents a semantic class and encodes Gaussian-distributed annotations at target locations. To address radar-specific characteristics, we incorporate Geometry Aware Conditioning and Temporal Consistency Regularization into the generative process. Experiments on the ROD2021 dataset demonstrate that signal reconstruction quality improves by \SI{3.6}{dB} in Peak Signal-to-Noise Ratio over baseline methods, while training with a combination of real and synthetic datasets improves overall mean Average Precision by 4.15% compared with conventional image-processing-based augmentation. These results indicate that our generative framework not only produces physically plausible and diverse radar spectrum but also substantially improves model generalization in downstream tasks.

Problem

• FMCW radar
• data scarcity
• dataset diversity
• environmental perception
• Range-Azimuth Maps

Innovation

• Generative Diffusion Model
• FMCW Radar
• Range-Azimuth Map
• Geometry Aware Conditioning
• Temporal Consistency Regularization