# AttentiveGRU: Recurrent Spatio-Temporal Modeling for Advanced Radar-Based BEV Object Detection

arXiv: https://arxiv.org/abs/2504.00559 · HTML: https://mlsanity.com/paper/2504.00559/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 35 references that are in the map.
- First posted on arXiv 1 Apr 2025.

## Builds on (most cited first, 23 of 23)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Feature Pyramid Networks for Object Detection](https://mlsanity.com/paper/1612.03144/) (2016)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [PointPillars: Fast Encoders for Object Detection from Point Clouds](https://mlsanity.com/paper/1812.05784/) (2018)
- [FCOS: Fully Convolutional One-Stage Object Detection](https://mlsanity.com/paper/1904.01355/) (2019)
- [Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting](https://mlsanity.com/paper/1506.04214/) (2015)
- [BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers](https://mlsanity.com/paper/2203.17270/) (2022)
- [Perceiver: General Perception with Iterative Attention](https://mlsanity.com/paper/2103.03206/) (2021)
- [Deformable ConvNets v2: More Deformable, Better Results](https://mlsanity.com/paper/1811.11168/) (2018)
- [PIXOR: Real-time 3D Object Detection from Point Clouds](https://mlsanity.com/paper/1902.06326/) (2019)
- [Dynamic Convolution: Attention over Convolution Kernels](https://mlsanity.com/paper/1912.03458/) (2019)
- [BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detection](https://mlsanity.com/paper/2203.17054/) (2022)
- [Delving Deeper into Convolutional Networks for Learning Video Representations](https://mlsanity.com/paper/1511.06432/) (2015)
- [RADIATE: A Radar Dataset for Automotive Perception in Bad Weather](https://mlsanity.com/paper/2010.09076/) (2020)
- [RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users](https://mlsanity.com/paper/2105.00363/) (2021)
- [Improved Orientation Estimation and Detection with Hybrid Object Detection Networks for Automotive Radar](https://mlsanity.com/paper/2205.02111/) (2022)
- [Exploiting Temporal Relations on Radar Perception for Autonomous Driving](https://mlsanity.com/paper/2204.01184/) (2022)
- [RaTrack: Moving Object Detection and Tracking with 4D Radar Point Cloud](https://mlsanity.com/paper/2309.09737/) (2023)
- [A recurrent CNN for online object detection on raw radar frames](https://mlsanity.com/paper/2212.11172/) (2022)
- [Improved Multi-Scale Grid Rendering of Point Clouds for Radar Object Detection Networks](https://mlsanity.com/paper/2305.15836/) (2023)
- [Quantification of Uncertainties in Deep Learning-based Environment Perception](https://mlsanity.com/paper/2306.03018/) (2023)
- [SIRA: Scalable Inter-frame Relation and Association for Radar Perception](https://mlsanity.com/paper/2411.02220/) (2024)
- [TimePillars: Temporally-Recurrent 3D LiDAR Object Detection](https://mlsanity.com/paper/2312.17260/) (2023)

## Built on by (most cited first, 3 of 3)

- [Graph Query Networks for Object Detection with Automotive Radar](https://mlsanity.com/paper/2511.15271/) (2025)
- [InterPruner: Interactive Structured Pruning via Taylor-Implicit Criterion and Language-Prior Modulator for Multimodal Object Detection](https://mlsanity.com/paper/2608.10724/) (2026)
- [If It Moves, Radar Knows: A Physics-Aware Radar Transformer for Class-Agnostic Moving-Object Detection](https://mlsanity.com/paper/2609.02289/) (2026)

## Abstract (arXiv)

Bird's-eye view (BEV) object detection has become important for advanced automotive 3D radar-based perception systems. However, the inherently sparse and non-deterministic nature of radar data limits the effectiveness of traditional single-frame BEV paradigms. In this paper, we addresses this limitation by introducing AttentiveGRU, a novel attention-based recurrent approach tailored for radar constraints, which extracts individualized spatio-temporal context for objects by dynamically identifying and fusing temporally correlated structures across present and memory states. By leveraging the consistency of object's latent representation over time, our approach exploits temporal relations to enrich feature representations for both stationary and moving objects, thereby enhancing detection performance and eliminating the need for externally providing or estimating any information about ego vehicle motion. Our experimental results on the public nuScenes dataset show a significant increase in mAP for the car category by 21% over the best radar-only submission. Further evaluations on an additional dataset demonstrate notable improvements in object detection capabilities, underscoring the applicability and effectiveness of our method.