Problem
Research questions and friction points this paper is trying to address.
MIMO radar
multi-modal fusion
autonomous vehicle perception
sparse array
end-to-end design
DeeperRadar: End-to-End MIMO Radar Design and Multi-Modal Fusion for Autonomous Vehicle Perception
🤖 AI Summary
This work addresses the limitations of conventional MIMO radar systems, whose fixed architectures are ill-suited for multimodal perception tasks and often lead to hardware redundancy and excessive cost. The authors propose a radar-centric, end-to-end framework that jointly optimizes sparse MIMO radar sampling patterns and multimodal 3D object detection by directly fusing raw radar ADC data, camera images, and LiDAR point clouds. The key innovation lies in introducing, for the first time, a task-driven, learnable MIMO array configuration: during training, an optimal receive antenna activation mask is learned under supervision from other modalities; at deployment, only this sparse mask is retained without requiring modifications to downstream models. Evaluated on the RADIal dataset, the method achieves comparable or superior performance to full-array baselines using significantly fewer antennas, substantially reducing system cost and integration complexity.
📝 Abstract
DeeperRadar is a radar-centric, sensor-stack-conditioned framework that co-designs radar sensing and multi-modal 3D detection for autonomous mobility by learning a sparse acquisition pattern end-to-end with the fusion model. A learnable MIMO design module is trained end-to-end within a fusion network that operates directly on raw radar ADC data together with camera images and LiDAR point clouds. During training, the design module is supervised by the other sensors, enabling the system to learn both which receiver antennas to activate and the effective number of them. At deployment, the design module is removed and replaced by the learned sparse subsampling mask, leaving the downstream model architecture unchanged. Evaluated on the RADIal dataset, DeeperRadar discovers sparse, task-aware radar configurations that match or exceed full-array baselines while using fewer receivers, potentially reducing radar cost and integration complexity. These results show that learned optimal MIMO radar design depends on the fusion stack and the downstream perception task.
Innovation
Methods, ideas, or system contributions that make the work stand out.
learnable MIMO radar
sparse acquisition
multi-modal fusion
end-to-end co-design
task-aware sensing
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
