Camera and millimeter-wave (MMW) radar fusion is essential for accurate and robust autonomous driving systems. With the advancement of radar technology, next-generation high-resolution automotive radar, i.e., 4D radar, has emerged. In addition to the target range, azimuth, and Doppler velocity measurements of traditional radar, 4D radar provides elevation measurement to create a denser “point cloud.” In this study, we propose a camera and 4D radar fusion network called RCFusion, which achieves multimodal feature fusion under a unified bird’s-eye view (BEV) space to accomplish 3D object detection tasks. In the camera stream, multi-scale feature maps are obtained by the image backbone and feature pyramid network; they are then converted into orthographic feature maps by an orthographic feature transform. Next, enhanced and fine-grained image BEV features are obtained via a designed shared attention encoder. Meanwhile, in the 4D radar stream, a newly designed component named Radar PillarNet efficiently encodes the radar features to generate radar pseudo-images, which are fed into the point cloud backbone to create radar BEV features. An interactive attention module is proposed for the fusion stage, which outputs a valid fusion of the two-modal BEV features. Finally, a generic detection head predicts the object classes and locations. The proposed RCFusion is validated on the TJ4DRadSet and View-of-Delft datasets. The experimental results and analysis show that the proposed method can effectively fuse camera and 4D radar features to achieve robust detection performance.

Object detection is a crucial task in autonomous driving. Currently, object-detection methods for autonomous driving systems are primarily based on information from cameras and lidar, which may experience interference from complex lighting or poor weather. At present, the four-dimensional (4D) (x, y, z, v) millimeter-wave radar can provide a denser point cloud to achieve 3D object-detection tasks that are difficult to complete with traditional millimeter-wave radar. Existing 3D object point-cloud detection algorithms are mostly based on 3D lidar, these methods are not necessarily applicable to millimeter-wave radars, which have sparser data and more noise and include velocity information. This study proposes a 3D object-detection framework based on a multi-frame 4D millimeter-wave radar point cloud. First, the ego vehicle velocity information is estimated by the millimeter-wave radar, and the relative velocity information of the millimeter-wave radar point cloud is compensated to the absolute velocity. Second, by matching between millimeter-wave radar frames, the multi-frame millimeter-wave radar point cloud is matched to the last frame. Finally, the object is detected by the proposed multi-frame millimeter-wave radar point cloud detection network. Experiments are performed using our newly recorded TJ4DRadSet dataset in a complex traffic environment. The results showed that the proposed object-detection framework outperformed the comparison methods based on the 3D mean average precision. The experimental results and methods can be used as the baseline for other multi-frame 4D millimeter-wave radar detection algorithms.

Automotive millimeter-wave (MMW) radar is essential in autonomous vehicles due to its robustness in all weather conditions. Traditional commercial automotive radars are limited by their resolution, which makes the object classification task difficult. Thus, the concept of a new generation of four-dimensional (4D) imaging radar was proposed. It has high azimuth and elevation resolution and contains Doppler information to produce a high-quality point cloud. In this paper, we propose an object classification network named Radar Transformer. The algorithm takes the attention mechanism as the core and adopts the combination of vector attention and scalar attention to make full use of the spatial information, Doppler information, and reflection intensity information of the radar point cloud to realize the deep fusion of local attention features and global attention features. We generated an imaging radar classification dataset and completed manual annotation. The experimental results show that our proposed method achieved an overall classification accuracy of 94.9%, which is more suitable for processing radar point clouds than the popular deep learning frameworks and shows promising performance.

4D millimeter-wave radar has gained attention as an emerging sensor for autonomous driving in recent years. However, existing 4D radar and camera fusion models often fail to fully exploit complementary information within each modality and lack deep cross-modal interactions. To address these issues, we propose a novel 4D radar and camera fusion method, named SGDet3D, for 3D object detection. Specifically, we first introduce a dual-branch fusion module that employs geometric depth completion and semantic radar PillarNet to comprehensively leverage geometric and semantic information within each modality. Then we introduce an object-oriented attention module that employs localization-aware cross-attention to facilitate deep interactions across modalites by allowing queries in bird's-eye view (BEV) to attend to interested image tokens. We validate our SGDet3D on the TJ4DRadSet and View-of-Delft (VoD) datasets. Experimental results demonstrate that SGDet3D effectively fuses 4D radar data and camera image and achieves state-of-the-art performance.

Four-dimensional (4D) radar–visual odometry (4DRVO) integrates complementary information from 4D radar and cameras, making it an attractive solution for achieving accurate and robust pose estimation. However, 4DRVO may exhibit significant tracking errors owing to three main factors: 1) sparsity of 4D radar point clouds; 2) inaccurate data association and insufficient feature interaction between the 4D radar and camera; and 3) disturbances caused by dynamic objects in the environment, affecting odometry estimation. In this paper, we present 4DRVO-Net, which is a method for 4D radar–visual odometry. This method leverages the feature pyramid, pose warping, and cost volume (PWC) network architecture to progressively estimate and refine poses. Specifically, we propose a multi-scale feature extraction network called Radar-PointNet++ that fully considers rich 4D radar point information, enabling fine-grained learning for sparse 4D radar point clouds. To effectively integrate the two modalities, we design an adaptive 4D radar–camera fusion module (A-RCFM) that automatically selects image features based on 4D radar point features, facilitating multi-scale cross-modal feature interaction and adaptive multi-modal feature fusion. In addition, we introduce a velocity-guided point-confidence estimation module to measure local motion patterns, reduce the influence of dynamic objects and outliers, and provide continuous updates during pose refinement. We demonstrate the excellent performance of our method and the effectiveness of each module design on both the VoD and in-house datasets. Our method outperforms all learning-based and geometry-based methods for most sequences in the VoD dataset. Furthermore, it has exhibited promising performance that closely approaches that of the 64-beam LiDAR odometry results of A-LOAM without mapping optimization.

Hand gesture recognition technology plays an important role in human-computer interaction and in-vehicle entertainment. Under in-vehicle conditions, it is a great challenge to design gesture recognition systems due to variable driving conditions, complex backgrounds, and diversified gestures. In this paper, we propose a gesture recognition system based on frequency-modulated continuous-wave (FMCW) radar and transformer for an in-vehicle environment. Firstly, the original range-Doppler maps (RDMs), range-azimuth maps (RAMs), and range-elevation maps (REMs) of the time sequence of each gesture are obtained by radar signal processing. Then we preprocess the obtained data frames by region of interest (ROI) extraction, vibration removal algorithm, background removal algorithm, and standardization. We propose a transformer-based radar gesture recognition network named RGTNet. It fully extracts and fuses the spatial-temporal information of radar feature maps to complete the classification of various gestures. The experimental results show that our method can better complete the eight gesture classification tasks in the in-vehicle environment. The recognition accuracy is 97.56%.