While cooperative perception can overcome the limitations of single-vehicle systems, the practical implementation of vehicle-to-vehicle and vehicle-to-infrastructure systems is often impeded by significant economic barriers. Aerial-ground cooperation (AGC), which pairs ground vehicles with drones, presents a more economically viable and rapidly deployable alternative. However, this emerging field has been held back by a critical lack of high-quality public datasets and benchmarks. To bridge this gap, we present Griffin, a comprehensive AGC 3D perception dataset, featuring over 250 dynamic scenes (37k+ frames). It incorporates varied drone altitudes (20-60m), diverse weather conditions, realistic drone dynamics via CARLA-AirSim co-simulation, and critical occlusion-aware 3D annotations. Accompanying the dataset is a unified benchmarking framework for cooperative detection and tracking, with protocols to evaluate communication efficiency, altitude adaptability, and robustness to communication latency, data loss and localization noise. By experiments through different cooperative paradigms, we demonstrate the effectiveness and limitations of current methods and provide crucial insights for future research.

Cooperative perception is critical for autonomous driving, overcoming the inherent limitations of a single vehicle, such as occlusions and constrained fields-of-view. However, current approaches sharing dense Bird's-Eye-View (BEV) features are constrained by quadratically-scaling communication costs and the lack of flexibility and interpretability for precise alignment across asynchronous or disparate viewpoints. While emerging sparse query-based methods offer an alternative, they often suffer from inadequate geometric representations, suboptimal fusion strategies, and training instability. In this paper, we propose SparseCoop, a fully sparse cooperative perception framework for 3D detection and tracking that completely discards intermediate BEV representations. Our framework features a trio of innovations: a kinematic grounded instance query that uses an explicit state vector with 3D geometry and velocity for precise spatio-temporal alignment; a coarse-to-fine aggregation module that effectively integrates information from both matched and unmatched instances; and a cooperative instance denoising task that provides stable, abundant supervision to accelerate and stabilize training. Experiments on V2X-Seq and Griffin datasets show SparseCoop achieves state-of-the-art performance. Notably, it delivers this performance with superior computational efficiency and a highly competitive transmission cost, while showing remarkable robustness to real-world challenges like communication latency.

Monocular 3D object detection is attractive for autonomous driving due to its low cost, but its performance degrades severely under adverse visual conditions that corrupts appearance cues and destabilizes depth-related representations. We propose Robo-DETR, a robustness-aware depth-guided transformer framework that combines minimal visual-conditioned input preconditioning with detector-internal depth reliability correction. Specifically, a lightweight Condition Recognition and Enhancement Block stabilizes low-level visual cues, while a Condition-aware Depth-guided Block refines depth logits via foreground-aware supervision and structured dynamic/region-level adjustment to suppress visual-induced depth noise. A dual-branch transformer decoder further promotes depth-appearance consistency through modality-specific cross-attention and fusion. Experiments on KITTI-C and the real-world TJ4DRadSet demonstrate consistent improvements over competitive monocular detectors under diverse degradations, and ablations validate the contribution of each component.

3D object detection plays a crucial role in autonomous driving, with Bird’s Eye View (BEV) becoming increasingly popular for its rich contextual information, ease of multi-modal fusion, and scalability. Despite its advantages, current BEV-based 3D detection methods still face significant challenges, including multi-modal fusion, communication bottlenecks, robustness under varying conditions, and safety concerns. This survey provides a systematic review of recent advancements in BEV perception, and meanwhile organizes a research based on these focal areas. It spans a broad range of perspectives, offering valuable insights for future perception research. Additionally, this survey explores the influence of emerging technologies, such as large language models and end-to-end frameworks on enhancing BEV perception capabilities, focusing on improving performance and robustness. Key future directions would include: 1) advancement from isolated vehicle perception to vehicle-to-everything (V2X) cooperative perception; 2) evolution from single-modal to integrated multi-modal fusion; 3) shift from simulated environments to real-world applications; and 4) transition from hierarchical perception frameworks to interpretable, end-to-end large-scale models.

Cooperatively utilizing both ego-vehicle and infrastructure sensor data via V2X communication has emerged as a promising approach for advanced autonomous driving. However, current research mainly focuses on improving individual modules, rather than taking end-to-end learning to optimize final planning performance, resulting in underutilized data potential. In this paper, we introduce UniV2X, a pioneering cooperative autonomous driving framework that seamlessly integrates all key driving modules across diverse views into a unified network. We propose a sparse-dense hybrid data transmission and fusion mechanism for effective vehicle-infrastructure cooperation, offering three advantages: 1) Effective for simultaneously enhancing agent perception, online mapping, and occupancy prediction, ultimately improving planning performance. 2) Transmission-friendly for practical and limited communication conditions. 3) Reliable data fusion with interpretability of this hybrid data. We implement UniV2X, as well as reproducing several benchmark methods, on the challenging DAIR-V2X, the real-world cooperative driving dataset. Experimental results demonstrate the effectiveness of UniV2X in significantly enhancing planning performance, as well as all intermediate output performance.

Cooperative perception aims to address the inherent limitations of single-vehicle autonomous driving systems through information exchange among multiple agents. Previous research has primarily focused on single-frame perception tasks. However, the more challenging cooperative sequential perception tasks, such as cooperative 3D multi-object tracking, have not been thoroughly investigated. Therefore, we propose CoopTrack, a fully instance-level end-to-end framework for cooperative tracking, featuring learnable instance association, which fundamentally differs from existing approaches. CoopTrack transmits sparse instance-level features that significantly enhance perception capabilities while maintaining low transmission costs. Furthermore, the framework comprises two key components: Multi-Dimensional Feature Extraction, and Cross-Agent Association and Aggregation, which collectively enable comprehensive instance representation with semantic and motion features, and adaptive cross-agent association and fusion based on a feature graph. Experiments on both the V2X-Seq and Griffin datasets demonstrate that CoopTrack achieves excellent performance. Specifically, it attains state-of-the-art results on V2X-Seq, with 39.0\% mAP and 32.8\% AMOTA. The project is available at https://github.com/zhongjiaru/CoopTrack.

BEV-based 3D perception with multi-frame images input is crucial for autonomous driving. However, current methods for temporal BEV perception fail to fully utilize long sequence features because of local fusion or high complexity. Recently, Mamba, a powerful temporal modeling network with linear complexity, has shown exceptional performance in various 2D vision tasks, but its application to 3D perception tasks remains unexplored. Therefore, this paper proposes a general BEV perception backbone named BEVMamba, which is the first work to leverage State Space Model for 3D perception. Built upon the BEVFormer, to adapt Mamba for 3D perception we first add Hybrid Positional Encoding to the BEV features, enabling the networks to be aware of their spatial-temporal position. In the Temporal SSM block, the proposed 3D Factorized Scan ensures that historical BEV features are enriched with global temporal-spatial information. Subsequently, the Spatial-Temporal Corridor Fusion aggregates all BEV features in a physically meaningful manner, achieving precise feature fusion. The reliable BEV features obtained by BEVMamba are used for various perception tasks, including 3D object detection and 3D occupancy prediction. Results on the nuScenes and Occ-3D nuScenes datasets show that BEVMamba outperforms its baseline BEVFormer in both dense and sparse perception tasks and demonstrates competitive performance compared to other methods, highlighting the potential of Mamba in 3D perception tasks. The code will be available at https://github.com/Liuxiaoaaa/bevmamba

The construction of local HD (High Definition) Map and Lane Network with onboard sensors is critical for autonomous vechicles and facilitates downstream tasks. In contrast to previous studies that treated building HD Map and Lane Network as two individual tasks, in this paper a unified BEV (Bird’s-Eye-View) perception framework is proposed with seperate decoders to realize two tasks simultaneously. In this paper, the gap between object detection and curve regression when using a DETR-like decoder is discussed and a curve region aware method is proposed to make up for the above gap. Specifically, a mechanism called Curve Region Aware Deformable Attention is designed with a bezier grid sampling module to guide the attention learning in bev features and structual loss regarding shapes of lanelines is also included. Moreover, a BEV spatial-temporal fusion method is introduced to better utilize historical features with minimal loss of spatial information. The results on NuScenes dataset show that our work has been close to or exceeded SOTAs (state-of-the-art) on both two tasks simultaneously.