检索结果-内蒙古大学图书馆

BEV transformer for visual 3d object detection applied with retentive mechanism

TRANSACTIONS OF THE INSTITUTE OF MEASUREMENT ANd CONTROL 2025年

作者： Pan, Jincheng Huang, Xiaoci Luo, Suyun Ma, Fang Shanghai Univ Engn Sci Shanghai 201620 Peoples R China

Three-dimensional (3d) vision perception tasks utilizing multiple cameras are pivotal for autonomous driving systems, encompassing both 3d object detection and map segmentation. We introduce a novel approach dubbed RetentiveBEV, leveraging Transformer to learn spatiotemporal features from Bird's Eye View (BEV) perspectives. These BEV representations form the foundational layer for further autonomous driving tasks. Succinctly, spatial features within regions of interest (ROIs) are harvested via spatial cross-attention, while temporal dynamics are integrated using temporal self-attention, enriching the BEV with historical data. Our spatial cross-attention is enhanced with a retentive mechanism, prioritizing information surrounding the focal points and enabling the decomposition of this attention mechanism to bolster computational efficiency. On the nuScenes data set test split, our approach achieves a nuScenes detection Score (NdS) score of 60.4%, without additional training data, which is an 8.7% improvement over the baseline (BEVFormer-base), and is close to the current state-of-the-art method SparseBEV, which gets NdS 65.7% as of August 2024. On the Val split of nuScenes, our method achieves the performance of 55.8 NdS while maintaining a real-time inference speed of 25.3 FPS, and we are currently working on further accelerating inference using TensorRT on the existing basis (the specification of mAP and NdS would be illustrated by equations (12) and (13)). The integration of the retentive mechanism notably boosts the precision and recall in 3d object detection while also expediting the inference process.

关键词： 3d object detection bird eye's view transformer deep learning retentive

来源：评论

学校读者我要写书评

暂无评论

Cross-Supervised LidAR-Camera Fusion for 3d object detection

引用

IEEE ACCESS 2025年 13卷 10447-10458页

作者： Zuo, Chao Jie Gu, Cao Yu Guo, Yi Kun Miao, Xiao dong Nanjing Tech Univ Sch Mech & Power Engn Nanjing 211889 Peoples R China

Fusing LidAR and camera information is essential for achieving accurate and reliable 3d object detection in autonomous driving systems. due to inherent differences between different modalities, seeking an efficient and accurate fusion method is of great importance. Recently, significant progress has been made in 3d object detection methods based on lift-splat-shot (LSS-based) approaches. However, inaccurate depth estimation and substantial semantic information loss remain significant factors limiting the accuracy of 3d detection. In this paper, we propose a cross-fusion framework under a dual spatial representation, by integrating information in different spatial representations, namely bird's-eye view (BEV) and camera view, and establishing soft links to fully utilize the information carried by different modalities. It consists of two important components, gated LidAR supervised BEV (GLS-BEV) and multi-attention cross fusion (MACF) modules. The former achieves accurate depth estimation by supervising the transformation of LidAR data with clear depth into the image space, constructing point cloud features in vehicle's perspective. The latter utilizes three sub-attention modules with different roles to achieve cross-modal interaction within the same space, effectively reducing semantic loss. On the nuScenes benchmark, our proposed method achieves outstanding 3d object detection results with 71.8 mAP and 74.2 NdS.

关键词： Feature extraction Three-dimensional displays Laser radar Point cloud compression Cameras object detection Semantics Accuracy Logic gates depth measurement 3d object detection LidAR-camera system multi-sensor fusion BEV

来源：评论

学校读者我要写书评

暂无评论

Point-Level Fusion and Channel Attention for 3d object detection in Autonomous driving

引用

SENSORS 2025年第4期25卷 1097页

作者： Shen, Juntao Fang, Zheng Huang, Jin Southwest Jiaotong Univ Sch Elect Engn Chengdu 611756 Peoples R China Sichuan Inst Land Sci & Technol Sichuan Ctr Satellite Applicat Technol Chengdu 610072 Peoples R China

As autonomous driving technology progresses, LidAR-based 3d object detection has emerged as a fundamental element of environmental perception systems. PointPillars transforms point cloud data into a two-dimensional pseudo-image and employs a 2d CNN for efficient and precise detection. Nevertheless, this approach encounters two primary challenges: (1) the sparsity and disorganization of raw point clouds hinder the model's capacity to capture local features, thus impacting detection accuracy;and (2) existing models struggle to detect small objects within complex environments, particularly regarding orientation estimation. To address these issues, we propose two enhancements: (1) point-level fusion of LidAR point clouds and RGB images, which integrates the semantic information of 2d images with the geometric features of 3d point clouds to improve model performance in intricate scenarios;(2) the incorporation of the Efficient Channel Attention mechanism to concentrate on essential features, particularly for small and sparse objects. Experimental results on the KITTI dataset indicate significant improvements, particularly in small object detection tasks, such as identifying pedestrians and cyclists. The enhanced model also demonstrates substantial gains in the Average Orientation Similarity (AOS) metric. These enhancements enhance the vehicle's ability to track and predict object trajectories in dynamic environments, critical for reliable recognition and decision-making.

关键词： autonomous driving 3d object detection multimodal fusion channel attention mechanism ECA

来源：评论

学校读者我要写书评

暂无评论

NCFdet: Enhanced point cloud features using the neural collapse phenomenon in multimodal fusion for 3d object detection

引用

JOURNAL OF COMPUTATIONAL dESIGN ANd ENGINEERING 2025年第1期12卷 300-311页

作者： Xu, Yaming Xu, Minglei Wang, Yan Li, Boliang Harbin Inst Technol Sch Astronaut Harbin 150000 Peoples R China

The accurate and effective detection of 3d objects represents a crucial component in the development of multi-sensor autonomous driving and robotics systems, particularly in the context of navigating complex urban environments. The complementary nature of image and point cloud data allows for the detection of objects with greater accuracy and robustness when both image and point cloud features are employed. At present, there is no optimal solution for the timing of multi-sensor fusion, particularly in the case of cross-modal data formats. In order to address these issues, we propose a multi-sensor object detection scheme based on Neural Collapse (NC) theory augmented point cloud, which we have designated NCFdet. In particular, we have incorporated an image pre-training model at the network layer where the NC occurs and designed a fusion module based on Transformer Attention. Furthermore, we investigate the interconnection between NC and transmodal transfer, and provide an explanation for the efficacy of the former based on the latter. The NCFdet system performs well for the detection of small targets in scenes due to the advantage of a compact image data format. The performance of the proposed NCFdet system was validated on the KITTI (Karlsruhe Institute of Technology and Toyota Technological Institute) dataset in comparison to existing methods. Furthermore, the object detection performance of the proposed framework was demonstrated on our visual rotation platform. The results show that the detection system achieves advanced fusion results. The code will be published after the paper is published.

关键词： 3d object detection multi-sensor systems neural collapse feature fusion

来源：评论

学校读者我要写书评

暂无评论

RagNet3d: Learning distinguishable representation for pooled grids in 3d object detection

引用

NEUROCOMPUTING 2025年 635卷

作者： Chen, Jiaxin Han, Yuehui Yan, Zhiqiang Qian, Jianjun Li, Jun Yang, Jian Nanjing Univ Sci & Technol PCA Lab Key Lab Intelligent Percept & Syst High Dimens Inf Minist Educ Nanjing Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Jiangsu Key Lab Image & Video Understanding Social Nanjing Peoples R China

3d object detection plays a crucial role in autonomous driving. Previous two-stage 3d detectors have developed Grid-based Region-of-Interest (RoI) Pooling techniques, such as RoI-Grid Pooling and Voxel Pooling, both of which quantify RoI into grids. However, the grids are usually ambiguous since parts of them pertain to multiple RoIs that identify a single object. To address this issue, we propose a RoI-Aware Grids Pooling Network (RagNet3d), which introduces RoI-View context to create distinguishable grid representations. Specifically, we present a RoI-View Prediction module that predicts RoI-View context via the guidance of the distance between RoIs and objects. Meanwhile, we propose a Couple-View Fusion module that propagates the probabilistic distribution, calculated from the RoI-View context, into the RoI-irrelevant grid features for further box refinement. Extensive experiments on KITTI and Waymo Open dataset show that our method achieves remarkable improvements against the baselines.

关键词： 3d object detection Grid based pooling RoI-View context

来源：评论

学校读者我要写书评

暂无评论

Corn pose estimation using 3d object detection and stereo images

引用

COMPUTERS ANd ELECTRONICS IN AGRICULTURE 2025年 231卷

作者： Gao, Yuliang Li, Zhen Hong, Qingqing Li, Bin Zhang, Lifeng Kyushu Inst Technol Grad Sch Engn Kitakyushu Fukuoka 8040015 Japan Nantong Univ Sch Elect Engn Nantong 226021 Jiangsu Peoples R China Yangzhou Univ Coll Artificial Intelligence Yangzhou 225012 Jiangsu Peoples R China

Corn is an important staple crop. Obtaining the pose and dimensions of corn is key for automating corn cultivation, mainly using robotic arms or similar devices to perform precise operations, such as pesticide spraying, measurement, or precise picking. In this study, the Stereo-Corn-Pose detection (SCPd) algorithm was proposed, which used three dimensional (3d) object detection to obtain the pose and dimensions of corn with stereo images. This algorithm includes pitch angle detection, which is absent in traditional 3d object detection. The SCPd algorithm consists of two models: the union box detection model, FCOS-Stereo, based on the anchor-free network FCOSNet, and the 3d bounding box regression model, Cross-Stereo-EfficientFormer. This regression model incorporates a cross-attention mechanism into EfficientFormer to extract and fuse features effectively. This work constructed a dataset comprising 2,700 samples for training and 300 samples for testing. In the test set, this work achieved a union bounding box mAP of 85.3%, representing a 3.5% improvement over the original FCOSNet model. It also achieved 88.3% AP3d and 85.6% APBEV for 3d bounding box regression, making increases of 5% and 4.7%, respectively, compared to the traditional 3d object detection method, IdA3d. The results indicate an accuracy of approximately 91% in detecting corn dimensions and pose. Therefore, the SCPd algorithm offers a novel framework for obtaining the 3d dimensions and pose of corn and promotes precision and smart agriculture for corn cultivation.

关键词： 3d object detection Stereo vision Corn Transformer FCOSNet

来源：评论

学校读者我要写书评

暂无评论

BEVFusion With dual Hard Instance Probing for Multimodal 3d object detection

引用

IEEE ACCESS 2025年 13卷 25546-25556页

作者： Kim, Taeho Kim, Joohee IIT Dept Elect & Comp Engn Chicago IL 60616 USA

False negatives (FN) in 3d object detection, which occur when small, distant, or hidden objects are missed, pose significant safety risks in autonomous driving systems. Recent multi-modal fusion methods have been proposed to enhance 3d object detection by combining the geometric accuracy of LidAR point clouds with the rich semantic features of camera images. However, few methods explicitly address false negatives, and many fail to effectively align and interact multimodal features during the fusion process. To address these challenges, we propose BEVFusion with dual Hard Instance Probing (BEVFusion-dHIP), a novel 3d object detection framework designed to systematically reduce false negatives. BEVFusion-dHIP incorporates Hard Instance Probing (HIP) into both LidAR BEV features and 3d position-aware image features, progressively refining the detection of challenging objects across multiple stages. Furthermore, we introduce a deformable Attention Fusion Network (dAFusionNet) to dynamically align and fuse LidAR and camera BEV features during the fusion process, effectively mitigating spatial misalignment and enhancing inter-modal feature interaction. Experimental results on the nuScenes dataset show that the proposed BEVFusion-dHIP outperforms state-of-the-art lidar and camera+lidar based 3d object detection models. For example, BEVFusion-dHIP achieves improvements of 3.0 and 3.2 in mAP and NdS, respectively, compared to the baseline model BEVFusion.

关键词： Three-dimensional displays Laser radar Cameras object detection Feature extraction Point cloud compression Hip Accuracy Semantics Fuses 3d object detection multi-modal transformer deformable attention deep learning

来源：评论

学校读者我要写书评

暂无评论

Bi-Att3ddet: Attention-Based Bi-directional Fusion for Multi-Modal 3d object detection

引用

SENSORS 2025年第3期25卷 658页

作者： Gao, Xu Zhao, Yaqian Wang, Yanan Shang, Jiandong Zhang, Chunmin Wu, Gang Zhengzhou Univ Sch Comp & Artificial Intelligence Zhengzhou 450001 Peoples R China Natl Supercomp Ctr Zhengzhou Zhengzhou 450001 Peoples R China Yutong Bus Co Ltd Zhengzhou 450000 Peoples R China

Currently, multi-modal 3d object detection methods have become a key area of research in the field of autonomous driving. Fusion is an essential factor affecting performance in multi-modal object detection. However, previous methods still suffer from the inability to effectively fuse features from LidAR and RGB images, resulting in a low utilization rate of complementary information between depth and semantic texture features. At the same time, existing methods may not adequately capture the structural information in Region of Interest (RoI) features when extracting them. Structural information plays a crucial role in RoI features. It encompasses the position, size, and orientation of objects, as well as the relative positions and spatial relationships between objects. Its absence can result in false or missed detections. To solve the above problems, we propose a multi-modal sensor fusion network, Bi-Att3ddet, which mainly consists of a Self-Attentive RoI Feature Extraction module (SARoIFE) and a Feature Bidirectional Interactive Fusion module (FBIF). Specifically, SARoIFE captures the relationship between different positions in RoI features to obtain high-quality RoI features through the self-attention mechanism. SARoIFE prepares for the fusion stage. FBIF performs bidirectional interaction between LidAR and pseudo RoI features to make full use of the complementary information. We perform comprehensive experiments on the KITTI dataset, and our method notably demonstrates a 1.55% improvement in the hard difficulty level and a 0.19% improvement in the mean Average Precision (mAP) metric on the test dataset.

关键词： multi-modal sensor fusion attention mechanism 3d object detection autonomous driving

来源：评论

学校读者我要写书评

暂无评论

SdPSNet: An Efficient 3d object detection Based on Spatial dynamic Pruning Sparse Convolution and Self-Attention Feature diffusion

引用

IEEE ACCESS 2025年 13卷 63703-63714页

作者： Wang, Meng Yu, Qianlei Liu, Haipeng Kunming Univ Sci & Technol Fac Informat Engn & Automat Kunming 650500 Peoples R China

In the field of autonomous driving, the technology of 3d object detection using LidAR point clouds has been widely implemented. However, existing detectors are faced with challenges due to redundant overhead when processing large amounts of background points in the depth perception field, and insufficient feature mapping in the central regions of large objects. To address these issues, this paper proposes a novel 3d object detection model, SdPSNet, which combines spatial dynamic pruning and self-attention feature diffusion to reduce data redundancy and improve the representation of central features. SdPSNet introduces a spatial dynamic pruning sparse convolution strategy to optimise the inference efficiency and data redundancy of the detector. Meanwhile, the self-attention feature diffusion module is used to effectively diffuse the features of the object edges to the central region, providing a richer and more accurate representation for modelling the centre of large objects. Extensive experimental results have been conducted on the Waymo Open, nuScenes and Argoverse2 datasets. On Argoverse2, SdPSNet achieves a 2.8% increase in mAP over the previous hybrid detector, HEdNet, while being $2.4\times $ faster. It outperforms the previously popular sparse detector FSdv2 by 2.3% in mAP and is $1.5\times $ faster. It also shows a 0.2% improvement in mAP and is $1.2\times $ faster than the current best-performing sparse detector, SAFdNet.

关键词： Feature extraction detectors Convolution Three-dimensional displays Point cloud compression Accuracy object detection Kernel Redundancy Solid modeling 3d object detection LidAR autonomous driving data redundancy feature diffusion

来源：评论

学校读者我要写书评

暂无评论

Efficient and robust multi-camera 3d object detection in bird-eye-view

引用

IMAGE ANd VISION COMPUTING 2025年 154卷

作者： Wang, Yuanlong Jiang, Hengtao Chen, Guanying Zhang, Tong Zhou, Jiaqing Qing, Zezheng Wang, Chunyan Zhao, Wanzhong Nanjing Univ Aeronaut & Astronaut Coll Energy & Power Engn Nanjing Jiangsu Peoples R China

Bird's-eye view (BEV) representations are increasingly used in autonomous driving perception due to their comprehensive, unobstructed vehicle surroundings. Compared to transformer or depth based methods, ray transformation based methods are more suitable for vehicle deployment and more efficient. However, these methods typically depend on accurate extrinsic camera parameters, making them vulnerable to performance degradation when calibration errors or installation changes occur. In this work, we follow ray transformation based methods and propose an extrinsic parameters free approach, which reduces reliance on accurate offline camera extrinsic calibration by using a neural network to predict extrinsic parameters online and can effectively improve the robustness of the model. In addition, we propose a multi-level and multi-scale image encoder to better encode image features and adopt a more intensive temporal fusion strategy. Our framework further mainly contains four important designs: (1) a multi-level and multi-scale image encoder, which can leverage multi-scale information on the inter-layer and the intra-layer for better performance, (2) ray-transformation with extrinsic parameters free approach, which can transfers image features to BEV space and lighten the impact of extrinsic disturbance on m-odel's detection performance, (3) an intensive temporal fusion strategy using motion information from five historical frames. (4) a high-performance BEV encoder that efficiently reduces the spatial dimensions of a voxel-based feature map and fuse the multi-scale and the multi-frame BEV features. Experiments on nuScenes show that our best model (R101@900 x 1600) realized competitive 41.7% mAP and 53.8% NdS on the validation set, which outperforming several state-of-the-art visual BEV models in 3d object detection.

关键词： Multi-camera 3d object detection Autonomous driving

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：