DRIFT: Dual-Representation Inter-Fusion Transformer for Automated Driving Perception with 4D Radar Point Clouds

• URL: http://arxiv.org/abs/2603.09695v2
• Date: Thu, 12 Mar 2026 06:06:36 GMT
• Title: DRIFT: Dual-Representation Inter-Fusion Transformer for Automated Driving Perception with 4D Radar Point Clouds
• Abstract summary: DRIFT is a model that captures and fuses both local and global contexts.<n>It is evaluated on the widely used View-of-Delft (VoD) dataset and a proprietary internal dataset.
• Score: 4.71547360356314
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 4D radars, which provide 3D point cloud data along with Doppler velocity, are attractive components of modern automated driving systems due to their low cost and robustness under adverse weather conditions. However, they provide a significantly lower point cloud density than LiDAR sensors. This makes it important to exploit not only local but also global contextual scene information. This paper proposes DRIFT, a model that effectively captures and fuses both local and global contexts through a dual-path architecture. The model incorporates a point path to aggregate fine-grained local features and a pillar path to encode coarse-grained global features. These two parallel paths are intertwined via novel feature-sharing layers at multiple stages, enabling full utilization of both representations. DRIFT is evaluated on the widely used View-of-Delft (VoD) dataset and a proprietary internal dataset. It outperforms the baselines on the tasks of object detection and/or free road estimation. For example, DRIFT achieves a mean average precision (mAP) of 52.6% (compared to, say, 45.4% of CenterPoint) on the VoD dataset.

Related papers

• FogDrive: A Multi-Modal Synthetic Driving Dataset for Perception under Graded Fog [0.0]
  We present FogDrive, a rigorously calibrated dataset bridging data-centric engineering and robust machine learning.<n>FogDrive contains 660 scenes across four synchronized cameras (RGB, depth, semantic segmentation), a LiDAR and semantic-LiDAR pair, and front radar.<n>Physically consistent fog is modeled independently on camera channels and LiDAR channels at three calibrated visibility.<n>A semantic-segmentation-based quality audit over 8k images validates annotations at 95.1% precision and over 99% recall for vehicles within 40m.
  arXiv  Detail & Related papers  (2026-07-18T06:53:56Z)
• MLF-4DRCNet: Multi-Level Fusion with 4D Radar and Camera for 3D Object Detection in Autonomous Driving [31.26862558777292]
  MLF-4DRCNet is a novel framework for 3D object detection via multi-level fusion of 4D radar and camera images.<n>It incorporates the point-, scene-, and proposal-level multi-modal information, enabling comprehensive feature representation.<n>It attains performance comparable to LiDAR-based models on the View-of-Delft dataset.
  arXiv  Detail & Related papers  (2025-09-23T04:02:28Z)
• Cross3DReg: Towards a Large-scale Real-world Cross-source Point Cloud Registration Benchmark [57.42211080221526]
  Cross-source point cloud registration, which aims to align point cloud data from different sensors, is a fundamental task in 3D vision.<n>The lack of publicly available large-scale real-world datasets for training the deep registration models, and the inherent differences in point clouds captured by multiple sensors pose challenges.<n>We construct Cross3DReg, the currently largest and real-world multi-modal cross-source point cloud registration dataset.<n>A visual-geometric attention guided matching module is proposed to enhance the consistency of cross-source point cloud features.
  arXiv  Detail & Related papers  (2025-09-08T09:01:13Z)
• ZFusion: An Effective Fuser of Camera and 4D Radar for 3D Object Perception in Autonomous Driving [7.037019489455008]
  We propose a 3D object detection method, termed ZFusion, which fuses 4D radar and vision modality.<n> FP-DDCA fuser packs Transformer blocks to interactively fuse multi-modal features at different scales.<n>Experiments show that ZFusion achieved the state-of-the-art mAP in the region of interest.
  arXiv  Detail & Related papers  (2025-04-04T13:29:32Z)
• MetaOcc: Spatio-Temporal Fusion of Surround-View 4D Radar and Camera for 3D Occupancy Prediction with Dual Training Strategies [12.485905108032146]
  This paper introduces MetaOcc, a novel multi-modal framework for omni-oriented 3D occupancy prediction.<n>To address the limitations of directly applying encoders to sparse radar data, we propose a Radar Height Self-Attention module.<n>To reduce reliance on expensive point cloud, we propose a pseudo-label generation pipeline based on an open-set segmentor.
  arXiv  Detail & Related papers  (2025-01-26T03:51:56Z)
• V2X-Radar: A Multi-modal Dataset with 4D Radar for Cooperative Perception [61.58737390490639]
  We present V2X-Radar, the first large-scale, real-world multi-modal dataset featuring 4D Radar.<n>The dataset consists of 20K LiDAR frames, 40K camera images, and 20K 4D Radar data, including 350K annotated boxes.<n>To support various research domains, we have established V2X-Radar-C for cooperative perception, V2X-Radar-I for roadside perception, and V2X-Radar-V for single-vehicle perception.
  arXiv  Detail & Related papers  (2024-11-17T04:59:00Z)
• Dual Radar: A Multi-modal Dataset with Dual 4D Radar for Autonomous Driving [22.633794566422687]
  We introduce a novel large-scale multi-modal dataset featuring, for the first time, two types of 4D radars captured simultaneously. Our dataset consists of 151 consecutive series, most of which last 20 seconds and contain 10,007 meticulously synchronized and annotated frames. We experimentally validate our dataset, providing valuable results for studying different types of 4D radars.
  arXiv  Detail & Related papers  (2023-10-11T15:41:52Z)
• DOLPHINS: Dataset for Collaborative Perception enabled Harmonious and Interconnected Self-driving [19.66714697653504]
  Vehicle-to-Everything (V2X) network has enabled collaborative perception in autonomous driving. The lack of datasets has severely blocked the development of collaborative perception algorithms. We release DOLPHINS: dataset for cOllaborative Perception enabled Harmonious and INterconnected Self-driving.
  arXiv  Detail & Related papers  (2022-07-15T17:07:07Z)
• Benchmarking the Robustness of LiDAR-Camera Fusion for 3D Object Detection [58.81316192862618]
  Two critical sensors for 3D perception in autonomous driving are the camera and the LiDAR. fusing these two modalities can significantly boost the performance of 3D perception models. We benchmark the state-of-the-art fusion methods for the first time.
  arXiv  Detail & Related papers  (2022-05-30T09:35:37Z)
• LiDAR-based 4D Panoptic Segmentation via Dynamic Shifting Network [56.71765153629892]
  We propose the Dynamic Shifting Network (DS-Net), which serves as an effective panoptic segmentation framework in the point cloud realm. Our proposed DS-Net achieves superior accuracies over current state-of-the-art methods in both tasks. We extend DS-Net to 4D panoptic LiDAR segmentation by the temporally unified instance clustering on aligned LiDAR frames.
  arXiv  Detail & Related papers  (2022-03-14T15:25:42Z)
• PC-DAN: Point Cloud based Deep Affinity Network for 3D Multi-Object Tracking (Accepted as an extended abstract in JRDB-ACT Workshop at CVPR21) [68.12101204123422]
  A point cloud is a dense compilation of spatial data in 3D coordinates. We propose a PointNet-based approach for 3D Multi-Object Tracking (MOT)
  arXiv  Detail & Related papers  (2021-06-03T05:36:39Z)
• Road Segmentation on low resolution Lidar point clouds for autonomous vehicles [3.6020689500145653]
  We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures.
  arXiv  Detail & Related papers  (2020-05-27T00:38:39Z)
• End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection [62.34374949726333]
  Pseudo-LiDAR (PL) has led to a drastic reduction in the accuracy gap between methods based on LiDAR sensors and those based on cheap stereo cameras. PL combines state-of-the-art deep neural networks for 3D depth estimation with those for 3D object detection by converting 2D depth map outputs to 3D point cloud inputs. We introduce a new framework based on differentiable Change of Representation (CoR) modules that allow the entire PL pipeline to be trained end-to-end.
  arXiv  Detail & Related papers  (2020-04-07T02:18:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.