Radar-Modulated Selection perturbs only the step size Δ and readout C parameters inside Mamba's selective scan with radar data while keeping other components image-only, yielding state-of-the-art depth estimation on nuScenes with up to 34% MAE reduction.
4d millimeter- wave radar in autonomous driving: A survey
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
representative citing papers
DCDA uses 4D radar-conditioned diffusion with dual critics to align degraded LiDAR features to a clean manifold, enabling generalization to unseen weather types and severities without paired data or labels.
RAVEN introduces a chirp-wise streaming radar perception network with MIMO-preserving encoders, learnable cross-antenna mixing, and early-exit to deliver competitive detection and BEV segmentation at reduced compute and latency.
AttenNKF augments InEKF with an attention-based neural compensator trained in latent space to correct foot-slip errors in legged robot state estimation.
PULSE stabilizes mmWave human pose estimation by screening Doppler motion prompts before injecting them into spatial magnitude reasoning.
ATN3D introduces density-aware early fusion with cross-modal gating, occupancy-gated neighborhood aggregation, evidence-conditioned channel self-attention, and a range-aware loss, reporting mAP gains of +3.55% (clear) and +8.41% (fog) on VoD, with larger relative gains for objects >30m.
Vision-radar fusion method uses depth-semantic alignment and affinity-guided fusion plus sparse completion to generate structured radar point clouds, improving object detection and tracking.
citing papers explorer
-
Selection, Not Fusion: Radar-Modulated State Space Models for Radar-Camera Depth Estimation
Radar-Modulated Selection perturbs only the step size Δ and readout C parameters inside Mamba's selective scan with radar data while keeping other components image-only, yielding state-of-the-art depth estimation on nuScenes with up to 34% MAE reduction.
-
Open-Weather Robust 3D Detection via Dual-Critic Diffusion Alignment
DCDA uses 4D radar-conditioned diffusion with dual critics to align degraded LiDAR features to a clean manifold, enabling generalization to unseen weather types and severities without paired data or labels.
-
RAVEN: Radar Adaptive Vision Encoders for Efficient Chirp-wise Object Detection and Segmentation
RAVEN introduces a chirp-wise streaming radar perception network with MIMO-preserving encoders, learnable cross-antenna mixing, and early-exit to deliver competitive detection and BEV segmentation at reduced compute and latency.
-
Attention-Based Neural-Augmented Kalman Filter for Legged Robot State Estimation
AttenNKF augments InEKF with an attention-based neural compensator trained in latent space to correct foot-slip errors in legged robot state estimation.
-
Doppler Prompting for Stable mmWave-based Human Pose Estimation
PULSE stabilizes mmWave human pose estimation by screening Doppler motion prompts before injecting them into spatial magnitude reasoning.
-
ATN3D: Density-Aware LiDAR-Radar Early 3D Object Detection Under Extreme Sparsity
ATN3D introduces density-aware early fusion with cross-modal gating, occupancy-gated neighborhood aggregation, evidence-conditioned channel self-attention, and a range-aware loss, reporting mAP gains of +3.55% (clear) and +8.41% (fog) on VoD, with larger relative gains for objects >30m.
-
Depth-Semantic Alignment and Affinity-Guided Fusion for Structured Radar Point Cloud Generation
Vision-radar fusion method uses depth-semantic alignment and affinity-guided fusion plus sparse completion to generate structured radar point clouds, improving object detection and tracking.
- Radar-Guided Polynomial Fitting for Metric Depth Estimation