Search

In camera-based bird's-eye view (BEV) 3D object detection, non-maximum suppression (NMS) plays a crucial role. However, traditional NMS methods become ineffective in BEV scenarios where the predicted bounding boxes of small object instances often have no overlapping areas. To address this issue, this paper proposes a BEV intersection over union (IoU) computation method based on relative position and absolute spatial information, referred to as B-IoU. Additionally, a BEV circular search method, called B-Grouping, is introduced to handle prediction boxes of varying scales. Utilizing these two methods, a novel NMS strategy called BEV-NMS is developed to handle the complex prediction boxes in BEV perspectives. This BEV-NMS strategy is implemented in several existing algorithms. Based on the results from the nuScenes validation set, there was an average increase of 7.9% in mAP when compared to the strategy without NMS. The NDS also showed an average increase of 7.9% under the same comparison. Furthermore, compared to the Scale-NMS strategy, the mAP increased by an average of 3.4%, and the NDS saw an average improvement of 3.1%. [ABSTRACT FROM AUTHOR]

As unmanned vehicle technology advances rapidly, obstacle recognition and target detection are crucial links, which directly affect the driving safety and efficiency of unmanned vehicles. In response to the inaccurate localization of small targets such as pedestrians in current object detection tasks and the problem of losing local features in the PointPillars, this paper proposes a three-dimensional object detection method based on improved PointPillars. Firstly, addressing the issue of lost spatial and local information in the PointPillars, the feature encoding part of the PointPillars is improved, and a new pillar feature enhancement extraction module, CSM-Module, is proposed. Channel encoding and spatial encoding are introduced in the new pillar feature enhancement extraction module, fully considering the spatial information and local detailed geometric information of each pillar, thereby enhancing the feature representation capability of each pillar. Secondly, based on the fusion of CSPDarknet and SENet, a new backbone network CSE-Net is designed in this paper, enabling the extraction of rich contextual semantic information and multi-scale global features, thereby enhancing the feature extraction capability. Our method achieves higher detection accuracy when validated on the KITTI dataset. Compared to the original network, the improved algorithm's average detection accuracy is increased by 3.42%, it shows that the method is reasonable and valuable. [ABSTRACT FROM AUTHOR]

Fueled by substantial advancements in deep learning, the domain of autonomous driving is swiftly advancing towards more robust and effective intelligent systems. One of the critical challenges in this field is achieving accurate 3D object detection, which is often hindered by data sparsity and occlusion. To address these issues, we propose a method centered around a multi-modal fusion strategy that leverages vehicle-road cooperation to enhance perception capabilities. Our approach integrates label information from roadside perception point clouds to harmonize and enrich the representation of image and LiDAR data. This comprehensive integration significantly improves detection accuracy by providing a fuller understanding of the surrounding environment. Rigorous evaluations of our proposed method on two benchmark datasets, KITTI and Waymo Open, demonstrate its superior performance, with our model achieving 87.52% 3D Average Precision (3D AP) and 93.71% Bird's Eye View Average Precision (BEV AP) on the KITTI v a l set. These results highlight the effectiveness of our method in detecting sparse and distant objects, contributing to the development of safer and more efficient autonomous driving solutions. [ABSTRACT FROM AUTHOR]

Copyright of Computer Measurement & Control is the property of Magazine Agency of Computer Measurement & Control and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Multimodal 3D object detection is an active research topic in the field of autonomous driving. Most existing methods utilize both camera and LiDAR modalities but fuse their features through simple and insufficient mechanisms. Additionally, these approaches lack reliable positional and temporal information due to their reliance on single-frame camera data. In this paper, a novel end-to-end framework for 3D object detection was proposed to solve these problems through spatial and temporal fusion. The spatial information of bird's-eye view (BEV) features is enhanced by integrating depth features from point clouds during the conversion of image features into 3D space. Moreover, positional and temporal information is augmented by aggregating multi-frame features. This framework is named as DMFusion, which consists of the following components: (i) a novel depth fusion view transform module (referred to as DFLSS), (ii) a simple and easily adjustable temporal fusion module based on 3D convolution (referred to as 3DMTF), and (iii) a LiDAR-temporal fusion module based on channel attention mechanism. On the nuScenes benchmark, DMFusion improves mAP by 1.42% and NDS by 1.26% compared with the baseline model, which demonstrates the effectiveness of our proposed method. The code will be released at https://github.com/lilkeker/DMFusion. [ABSTRACT FROM AUTHOR]

Point cloud-based 3D small object detection is crucial for autonomous driving and smart ships. The current 3D object detection mainly relies on object global features derived from 3D and 2D convolutional networks, inevitably leading to the loss of substantial object detail information. As large objects contain enough points to obtain sufficient global features, they are easy to identify. In contrast, small objects contain fewer points and their global features are weak, resulting in false classifications and inaccurate location estimates. Therefore, it is necessary to take into account the local geometric relation features of the object to develop adequate discriminative features. Moreover, the current two-stage 3D object detection speed is relatively slow due to the complex refinement structure, which is adverse to real-time detection. In this paper, an efficient 3D small object detection network with two novel modules is proposed. Firstly, the Geometric relation-based Feature Aggregation (GFA) module is designed to improve small object detection performance. This module flexibly aggregates the features of voxels and original points near the key points, for key points to aggregate more local discriminate features of objects, which is conducive to small object detection. Subsequently, the Key point Feature Abstraction (KFA) module is designed to improve the speed of small object detection, through which object global features can be rapidly obtained and the detection performance can be enhanced. Experimental results show that this method achieves state-of-the-art small object detection performance on both the KITTI dataset and the River Cargo Ship dataset. [ABSTRACT FROM AUTHOR]

Quadrupedal robots are confronted with the intricate challenge of navigating dynamic environments fraught with diverse and unpredictable scenarios. Effectively identifying and responding to obstacles is paramount for ensuring safe and reliable navigation. This paper introduces a pioneering method for 3D object detection, termed viewpoint feature histograms, which leverages the established paradigm of 2D detection in projection. By translating 2D bounding boxes into 3D object proposals, this approach not only enables the reuse of existing 2D detectors but also significantly increases the performance with less computation required, allowing for real-time detection. Our method is versatile, targeting both bird's eye view objects (e.g., cars) and frontal view objects (e.g., pedestrians), accommodating various types of 2D object detectors. We showcase the efficacy of our approach through the integration of YOLO3D, utilizing LiDAR point clouds on the KITTI dataset, to achieve real-time efficiency aligned with the demands of autonomous vehicle navigation. Our model selection process, tailored to the specific needs of quadrupedal robots, emphasizes considerations such as model complexity, inference speed, and customization flexibility, achieving an accuracy of up to 99.93%. This research represents a significant advancement in enabling quadrupedal robots to navigate complex and dynamic environments with heightened precision and safety. [ABSTRACT FROM AUTHOR]

Accurate recognition and localization of 3D objects is a fundamental research problem in 3D computer vision. Benefiting from transformation-free point cloud processing and flexible receptive fields, point-based methods have become accurate in 3D point cloud modeling, but still fall behind voxel-based competitors in 3D detection. We observe that the set abstraction module, commonly utilized by point-based methods for downsampling points, tends to retain excessive irrelevant background information, thus hindering the effective learning of features for object detection tasks. To address this issue, we propose MSSA, a Multi-representation Semantics-augmented Set Abstraction for 3D object detection. Specifically, we first design a backbone network to encode different representation features of point clouds, which extracts point-wise features through PointNet to preserve fine-grained geometric structure features, and adopts VoxelNet to extract voxel features and BEV features to enhance the semantic features of key points. Second, to efficiently fuse different representation features of keypoints, we propose a Point feature-guided Voxel feature and BEV feature fusion (PVB-Fusion) module to adaptively fuse multi-representation features and remove noise. At last, a novel Multi-representation Semantic-guided Farthest Point Sampling (MS-FPS) algorithm is designed to help set abstraction modules progressively downsample point clouds, thereby improving instance recall and detection performance with more important foreground points. We evaluate MSSA on the widely used KITTI dataset and the more challenging nuScenes dataset. Experimental results show that compared to PointRCNN, our method improves the AP of "moderate" level for three classes of objects by 7.02%, 6.76%, and 5.44%, respectively. Compared to the advanced point-voxel-based method PV-RCNN, our method improves the AP of "moderate" level by 1.23%, 2.84%, and 0.55% for the three classes, respectively. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In order to accurately identify the target obstacle in the traffic scene, considering the complexity of the real road scene and the importance of road safety, taking the sparsely embedded convolutional detection( SECOND) model as the basic model, the point cloud representation ability was enhanced by using the self-attention mechanism to obtain global semantic information, and the region of interest (RoI) detection head was used to optimize the 3D suggestion box generated by the candidate region to improve its detection accuracy. Then, two-stage 3D object detection method based on self-attention mechanism named SAR-SECOND was proposed. The results show that compared with the existing advanced 3D object detection methods, SAR-SECOND overall detection accuracy of car is 82. 28%, the overall accuracy of pedestrians is 51. 45%, and the overall accuracy of the rider is 72. 41% . It can be seen that the effectiveness of this method is effective. [ABSTRACT FROM AUTHOR]

The development of self-driving or autonomous vehicles has led to significant advancements in 3D object detection technologies, which are critical for the safety and efficiency of autonomous driving. Despite recent advances, several challenges remain in sensor integration, handling sparse and noisy data, and ensuring reliable performance across diverse environmental conditions. This paper comprehensively surveys state-of-the-art 3D object detection techniques for autonomous vehicles, emphasizing the importance of multi-sensor fusion techniques and advanced deep learning models. Furthermore, we present key areas for future research, including enhancing sensor fusion algorithms, improving computational efficiency, and addressing ethical, security, and privacy concerns. The integration of these technologies into real-world applications for autonomous driving is presented by highlighting potential benefits and limitations. We also present a side-by-side comparison of different techniques in a tabular form. Through a comprehensive review, this paper aims to provide insights into the future directions of 3D object detection and its impact on the evolution of autonomous driving. [ABSTRACT FROM AUTHOR]

Current state-of-the-art (SOTA) LiDAR-only detectors perform well for 3D object detection tasks, but point cloud data are typically sparse and lacks semantic information. Detailed semantic information obtained from camera images can be added with existing LiDAR-based detectors to create a robust 3D detection pipeline. With two different data types, a major challenge in developing multi-modal sensor fusion networks is to achieve effective data fusion while managing computational resources. With separate 2D and 3D feature extraction backbones, feature fusion can become more challenging as these modes generate different gradients, leading to gradient conflicts and suboptimal convergence during network optimization. To this end, we propose a 3D object detection method, Attention-Enabled Point Fusion (AEPF). AEPF uses images and voxelized point cloud data as inputs and estimates the 3D bounding boxes of object locations as outputs. An attention mechanism is introduced to an existing feature fusion strategy to improve 3D detection accuracy and two variants are proposed. These two variants, AEPF-Small and AEPF-Large, address different needs. AEPF-Small, with a lightweight attention module and fewer parameters, offers fast inference. AEPF-Large, with a more complex attention module and increased parameters, provides higher accuracy than baseline models. Experimental results on the KITTI validation set show that AEPF-Small maintains SOTA 3D detection accuracy while inferencing at higher speeds. AEPF-Large achieves mean average precision scores of 91.13, 79.06, and 76.15 for the car class's easy, medium, and hard targets, respectively, in the KITTI validation set. Results from ablation experiments are also presented to support the choice of model architecture. [ABSTRACT FROM AUTHOR]

Aiming at the limitations of the existing railway foreign object detection methods based on two-dimensional (2D) images, such as short detection distance, strong influence of environment and lack of distance information, we propose Rail-PillarNet, a three-dimensional (3D) LIDAR (Light Detection and Ranging) railway foreign object detection method based on the improvement of PointPillars. Firstly, the parallel attention pillar encoder (PAPE) is designed to fully extract the features of the pillars and alleviate the problem of local fine-grained information loss in PointPillars pillars encoder. Secondly, a fine backbone network is designed to improve the feature extraction capability of the network by combining the coding characteristics of LIDAR point cloud feature and residual structure. Finally, the initial weight parameters of the model were optimised by the transfer learning training method to further improve accuracy. The experimental results on the OSDaR23 dataset show that the average accuracy of Rail-PillarNet reaches 58.51%, which is higher than most mainstream models, and the number of parameters is 5.49 M. Compared with PointPillars, the accuracy of each target is improved by 10.94%, 3.53%, 16.96% and 19.90%, respectively, and the number of parameters only increases by 0.64 M, which achieves a balance between the number of parameters and accuracy. [ABSTRACT FROM AUTHOR]

Automatic 3D object detection using monocular cameras presents significant challenges in the context of autonomous driving. Precise labeling of 3D object scales requires accurate spatial information, which is difficult to obtain from a single image due to the inherent lack of depth information in monocular images, compared to LiDAR data. In this paper, we propose a novel approach to address this issue by enhancing deep neural networks with depth information for monocular 3D object detection. The proposed method comprises three key components: 1)Feature Enhancement Pyramid Module: We extend the conventional Feature Pyramid Networks (FPN) by introducing a feature enhancement pyramid network. This module fuses feature maps from the original pyramid and captures contextual correlations across multiple scales. To increase the connectivity between low-level and high-level features, additional pathways are incorporated. 2)Auxiliary Dense Depth Estimator: We introduce an auxiliary dense depth estimator that generates dense depth maps to enhance the spatial perception capabilities of the deep network model without adding computational burden. 3)Augmented Center Depth Regression: To aid center depth estimation, we employ additional bounding box vertex depth regression based on geometry. Our experimental results demonstrate the superiority of the proposed technique over existing competitive methods reported in the literature. The approach showcases remarkable performance improvements in monocular 3D object detection, making it a promising solution for autonomous driving applications. [ABSTRACT FROM AUTHOR]

Cameras and LiDAR are important sensors in autonomous driving systems that can provide complementary information to each other. However, most LiDAR-only methods outperform the fusion method on the main benchmark datasets. Current studies attribute the reasons for this to misalignment of views and difficulty in matching heterogeneous features. Specially, using the single-stage fusion method, it is difficult to fully fuse the features of the image and point cloud. In this work, we propose a 3D object detection network based on the multi-layer and multi-modal fusion (3DMMF) method. 3DMMF works by painting and encoding the point cloud in the frustum proposed by the 2D object detection network. Then, the painted point cloud is fed to the LiDAR-only object detection network, which has expanded channels and a self-attention mechanism module. Finally, the camera-LiDAR object candidates fusion for 3D object detection(CLOCs) method is used to match the geometric direction features and category semantic features of the 2D and 3D detection results. Experiments on the KITTI dataset (a public dataset) show that this fusion method has a significant improvement over the baseline of the LiDAR-only method, with an average mAP improvement of 6.3%. [ABSTRACT FROM AUTHOR]

The fusion of visual and millimeter-wave radar data has emerged as a prominent solution for precise 3D object detection. This paper focuses on the fusion of visual and mmWave radar information and presents an enhanced fusion method called CenRadfusion. This method represents an evolution and improvement over the classic CenterFusion network by leveraging the fused features from mmWave radar and camera data to achieve accurate 3D object detection. The key features of this method are as follows:To ensure the integrity of the fusion architecture, mmWave radar point clouds are initially projected onto the image plane and added as an additional channel to the input of the CenterNet image detection network. This process forms preliminary 3D detection boxes.Subsequently, mmWave radar point clouds are subjected to density-based clustering, which results in the acquisition of labels and the elimination of irrelevant point clouds and white noise. This step enhances data quality and the reliability of object detection.Finally, an attention module, known as the Squeeze-and-Excitation Networks, is incorporated to weight each feature channel, thereby enhancing the importance of crucial features in the network.Experimental results demonstrate that compared to the original CenterFusion algorithm, the detection Average Precision (AP) values for cars, trucks, and motorcycles have improved by 7.8%, 5.5%, and 5.4%, respectively. [ABSTRACT FROM AUTHOR]

LiDAR sensors have been shown to generate data with various common corruptions, which seriously affect their applications in 3D vision tasks, particularly object detection. At the same time, it has been demonstrated that traditional defense strategies, including adversarial training, are prone to suffering from gradient confusion during training. Moreover, they can only improve their robustness against specific types of data corruption. In this work, we propose LiDARPure, which leverages the powerful generation ability of diffusion models to purify corruption in the LiDAR scene data. By dividing the entire scene into voxels to facilitate the processes of diffusion and reverse diffusion, LiDARPure overcomes challenges induced from adversarial training, such as sparse point clouds in large-scale LiDAR data and gradient confusion. In addition, we utilize the latent geometric features of a scene as a condition to assist the generation of diffusion models. Detailed experiments show that LiDARPure can effectively purify 19 common types of LiDAR data corruption. Further evaluation results demonstrate that it can improve the average precision of 3D object detectors to an extent of 20% in the face of data corruption, much higher than existing defence strategies. [ABSTRACT FROM AUTHOR]

Performing tasks in agriculture, such as fruit monitoring or harvesting, requires perceiving the objects' spatial position. RGB-D cameras are limited under open-field environments due to lightning interferences. So, in this study, we state to answer the research question: "How can we use and control monocular sensors to perceive objects' position in the 3D task space?" Towards this aim, we approached histogram filters (Bayesian discrete filters) to estimate the position of tomatoes in the tomato plant through the algorithm MonoVisual3DFilter. Two kernel filters were studied: the square kernel and the Gaussian kernel. The implemented algorithm was essayed in simulation, with and without Gaussian noise and random noise, and in a testbed at laboratory conditions. The algorithm reported a mean absolute error lower than 10 mm in simulation and 20 mm in the testbed at laboratory conditions with an assessing distance of about 0.5 m. So, the results are viable for real environments and should be improved at closer distances. [ABSTRACT FROM AUTHOR]

The construction of virtual indoor spaces is crucial for the development of metaverses, virtual production, and other 3D content domains. Traditional methods for creating these spaces are often cost-prohibitive and labor-intensive. To address these challenges, we present a pipeline for generating digital twins of real indoor environments from RGB-D camera-scanned data. Our pipeline synergizes space structure estimation, 3D object detection, and the inpainting of missing areas, utilizing deep learning technologies to automate the creation process. Specifically, we apply deep learning models for object recognition and area inpainting, significantly enhancing the accuracy and efficiency of virtual space construction. Our approach minimizes manual labor and reduces costs, paving the way for the creation of metaverse spaces that closely mimic real-world environments. Experimental results demonstrate the effectiveness of our deep learning applications in overcoming traditional obstacles in digital twin creation, offering high-fidelity digital replicas of indoor spaces. This advancement opens for immersive and realistic virtual content creation, showcasing the potential of deep learning in the field of virtual space construction. [ABSTRACT FROM AUTHOR]

While cycling presents environmental benefits and promotes a healthy lifestyle, the risks associated with overtaking maneuvers by motorized vehicles represent a significant barrier for many potential cyclists. A large-scale analysis of overtaking maneuvers could inform traffic researchers and city planners how to reduce these risks by better understanding these maneuvers. Drawing from the fields of sensor-based cycling research and from LiDAR-based traffic data sets, this paper provides a step towards addressing these safety concerns by introducing the Salzburg Bicycle 3d (SaBi3d) data set, which consists of LiDAR point clouds capturing car-to-bicycle overtaking maneuvers. The data set, collected using a LiDAR-equipped bicycle, facilitates the detailed analysis of a large quantity of overtaking maneuvers without the need for manual annotation through enabling automatic labeling by a neural network. Additionally, a benchmark result for 3D object detection using a competitive neural network is provided as a baseline for future research. The SaBi3d data set is structured identically to the nuScenes data set, and therefore offers compatibility with numerous existing object detection systems. This work provides valuable resources for future researchers to better understand cycling infrastructure and mitigate risks, thus promoting cycling as a viable mode of transportation. Dataset: https://osf.io/k7cg9 (accessed on 18 July 2024). Dataset License: CC-By Attribution 4.0 International. [ABSTRACT FROM AUTHOR]

Despite the recent advancements in 3D object detection, the conventional 3D point cloud object detection algorithms have been found to exhibit limited accuracy for the detection of small objects. To address the challenge of poor detection of small-scale objects, this paper adopts the PointPillars algorithm as the baseline model and proposes a two-stage 3D target detection approach. As a cutting-edge solution, point cloud processing is performed using Transformer models. Additionally, a redefined attention mechanism is introduced to further enhance the detection capabilities of the algorithm. In the first stage, the algorithm uses PointPillars as the baseline model. The central concept of this algorithm is to transform the point cloud space into equal-sized columns. During the feature extraction stage, when the features from all cylinders are transformed into pseudo-images, the proposed algorithm incorporates attention mechanisms adapted from the Squeeze-and-Excitation (SE) method to emphasize and suppress feature information. Furthermore, the 2D convolution of the traditional backbone network is replaced by dynamic convolution. Concurrently, the addition of the attention mechanism further improves the feature representation ability of the network. In the second phase, the candidate frames generated in the first phase are refined using a Transformer-based approach. The proposed algorithm applies channel weighting in the decoder to enhance channel information, leading to improved detection accuracy and reduced false detections. The encoder constructs the initial point features from the candidate frames for encoding. Meanwhile, the decoder applies channel weighting to enhance the channel information, thereby improving the detection accuracy and reducing false detections. In the KITTI dataset, the experimental results verify the effectiveness of this method in small objects detection. Experimental results show that the proposed method significantly improves the detection capability of small objects compared with the baseline PointPillars. In concrete terms, in the moderate difficulty detection category, cars, pedestrians, and cyclists average precision (AP) values increased by 5.30%, 8.1%, and 10.6%, respectively. Moreover, the proposed method surpasses existing mainstream approaches in the cyclist category. [ABSTRACT FROM AUTHOR]

Multi-modal sensors are the key to ensuring the robust and accurate operation of autonomous driving systems, where LiDAR and cameras are important on-board sensors. However, current fusion methods face challenges due to inconsistent multi-sensor data representations and the misalignment of dynamic scenes. Specifically, current fusion methods either explicitly correlate multi-sensor data features by calibrating parameters, ignoring the feature blurring problems caused by misalignment, or find correlated features between multi-sensor data through global attention, causing rapidly escalating computational costs. On this basis, we propose a transformer-based end-to-end multi-sensor fusion framework named the adaptive fusion transformer (AFTR). The proposed AFTR consists of the adaptive spatial cross-attention (ASCA) mechanism and the spatial temporal self-attention (STSA) mechanism. Specifically, ASCA adaptively associates and interacts with multi-sensor data features in 3D space through learnable local attention, alleviating the problem of the misalignment of geometric information and reducing computational costs, and STSA interacts with cross-temporal information using learnable offsets in deformable attention, mitigating displacements due to dynamic scenes. We show through numerous experiments that the AFTR obtains SOTA performance in the nuScenes 3D object detection task (74.9% NDS and 73.2% mAP) and demonstrates strong robustness to misalignment (only a 0.2% NDS drop with slight noise). At the same time, we demonstrate the effectiveness of the AFTR components through ablation studies. In summary, the proposed AFTR is an accurate, efficient, and robust multi-sensor data fusion framework.

Abstract Three‐dimensional object detection based on the fusion of 2D image data and 3D point clouds has become a research hotspot in the field of 3D scene understanding. However, different sensor data have discrepancies in spatial position, scale, and alignment, which severely impact detection performance. Inappropriate fusion methods can lead to the loss and interference of valuable information. Therefore, we propose the High‐Order Multi‐Level Attention Fusion Network (HMAF‐Net), which takes camera images and voxelized point clouds as inputs for 3D object detection. To enhance the expressive power between different modality features, we introduce a high‐order feature fusion module that performs multi‐level convolution operations on the element‐wise summed features. By incorporating filtering and non‐linear activation, we extract deep semantic information from the fused multi‐modal features. To maximize the effectiveness of the fused salient feature information, we introduce an attention mechanism that dynamically evaluates the importance of pooled features at each level, enabling adaptive weighted fusion of significant and secondary features. To validate the effectiveness of HMAF‐Net, we conduct experiments on the KITTI dataset. In the “Car,” “Pedestrian,” and “Cyclist” categories, HMAF‐Net achieves mAP performances of 81.78%, 60.09%, and 63.91%, respectively, demonstrating more stable performance compared to other multi‐modal methods. Furthermore, we further evaluate the framework's effectiveness and generalization capability through the KITTI benchmark test, and compare its performance with other published detection methods on the 3D detection benchmark and BEV detection benchmark for the “Car” category, showing excellent results. The code and model will be made available on https://github.com/baowenzhang/High‐order‐Multilayer‐Attention‐Fusion‐Network‐for‐3D‐Object‐Detection.

The development of self-driving or autonomous vehicles has led to significant advancements in 3D object detection technologies, which are critical for the safety and efficiency of autonomous driving. Despite recent advances, several challenges remain in sensor integration, handling sparse and noisy data, and ensuring reliable performance across diverse environmental conditions. This paper comprehensively surveys state-of-the-art 3D object detection techniques for autonomous vehicles, emphasizing the importance of multi-sensor fusion techniques and advanced deep learning models. Furthermore, we present key areas for future research, including enhancing sensor fusion algorithms, improving computational efficiency, and addressing ethical, security, and privacy concerns. The integration of these technologies into real-world applications for autonomous driving is presented by highlighting potential benefits and limitations. We also present a side-by-side comparison of different techniques in a tabular form. Through a comprehensive review, this paper aims to provide insights into the future directions of 3D object detection and its impact on the evolution of autonomous driving.

Abstract In the wave of research on autonomous driving, 3D object detection from the Bird’s Eye View (BEV) perspective has emerged as a pivotal area of focus. The essence of this challenge is the effective fusion of camera and LiDAR data into the BEV. Current approaches predominantly train and predict within the front view and Cartesian coordinate system, often overlooking the inherent structural and operational differences between cameras and LiDAR sensors. This paper introduces CL-FusionBEV, an innovative 3D object detection methodology tailored for sensor data fusion in the BEV perspective. Our approach initiates with a view transformation, facilitated by an implicit learning module that transitions the camera’s perspective to the BEV space, thereby aligning the prediction module. Subsequently, to achieve modal fusion within the BEV framework, we employ voxelization to convert the LiDAR point cloud into BEV space, thereby generating LiDAR BEV spatial features. Moreover, to integrate the BEV spatial features from both camera and LiDAR, we have developed a multi-modal cross-attention mechanism and an implicit multi-modal fusion network, designed to enhance the synergy and application of dual-modal data. To counteract potential deficiencies in global reasoning and feature interaction arising from multi-modal cross-attention, we propose a BEV self-attention mechanism that facilitates comprehensive global feature operations. Our methodology has undergone rigorous evaluation on a substantial dataset within the autonomous driving domain, the nuScenes dataset. The outcomes demonstrate that our method achieves a mean Average Precision (mAP) of 73.3% and a nuScenes Detection Score (NDS) of 75.5%, particularly excelling in the detection of cars and pedestrians with high accuracies of 89% and 90.7%, respectively. Additionally, CL-FusionBEV exhibits superior performance in identifying occluded and distant objects, surpassing existing comparative methods.

Abstract The rapid growth on the amount of generated 3D data, particularly in the form of Light Detection And Ranging (LiDAR) point clouds (PCs), poses very significant challenges in terms of data storage, transmission, and processing. Point cloud (PC) representation of 3D visual information has shown to be a very flexible format with many applications ranging from multimedia immersive communication to machine vision tasks in the robotics and autonomous driving domains. In this paper, we investigate the performance of four reference 3D object detection techniques, when the input PCs are compressed with varying levels of degradation. Compression is performed using two MPEG standard coders based on 2D projections and octree decomposition, as well as two coding methods based on Deep Learning (DL). For the DL coding methods, we used a Joint Photographic Experts Group (JPEG) reference PC coder, that we adapted to accept LiDAR PCs in both Cartesian and cylindrical coordinate systems. The detection performance of the four reference 3D object detection methods was evaluated using both pre-trained models and models specifically trained using degraded PCs reconstructed from compressed representations. It is shown that LiDAR PCs can be compressed down to 6 bits per point with no significant degradation on the object detection precision. Furthermore, employing specifically trained detection models improves the detection capabilities even at compression rates as low as 2 bits per point. These results show that LiDAR PCs can be coded to enable efficient storage and transmission, without significant object detection performance loss.

3D object detection is a challenging and promising task for autonomous driving and robotics, benefiting significantly from multi-sensor fusion, such as LiDAR and cameras. Conventional methods for sensor fusion rely on a projection matrix to align the features from LiDAR and cameras. However, these methods often suffer from inadequate flexibility and robustness, leading to lower alignment accuracy under complex environmental conditions. Addressing these challenges, in this paper, we propose a novel Bidirectional Attention Fusion module, named BAFusion, which effectively fuses the information from LiDAR and cameras using cross-attention. Unlike the conventional methods, our BAFusion module can adaptively learn the cross-modal attention weights, making the approach more flexible and robust. Moreover, drawing inspiration from advanced attention optimization techniques in 2D vision, we developed the Cross Focused Linear Attention Fusion Layer (CFLAF Layer) and integrated it into our BAFusion pipeline. This layer optimizes the computational complexity of attention mechanisms and facilitates advanced interactions between image and point cloud data, showcasing a novel approach to addressing the challenges of cross-modal attention calculations. We evaluated our method on the KITTI dataset using various baseline networks, such as PointPillars, SECOND, and Part-A2, and demonstrated consistent improvements in 3D object detection performance over these baselines, especially for smaller objects like cyclists and pedestrians. Our approach achieves competitive results on the KITTI benchmark. [ABSTRACT FROM AUTHOR]

Recently, the growing demand for autonomous driving in the industry has led to a lot of interest in 3D object detection, resulting in many excellent 3D object detection algorithms. However, most 3D object detectors focus only on a single set of LiDAR points, ignoring their potential ability to improve performance by leveraging the information provided by the consecutive set of LIDAR points. In this paper, we propose a novel 3D object detection method called temporal motion-aware 3D object detection (TM3DOD), which utilizes temporal LiDAR data. In the proposed TM3DOD method, we aggregate LiDAR voxels over time and the current BEV features by generating motion features using consecutive BEV feature maps. First, we present the temporal voxel encoder (TVE), which generates voxel representations by capturing the temporal relationships among the point sets within a voxel. Next, we design a motion-aware feature aggregation network (MFANet), which aims to enhance the current BEV feature representation by quantifying the temporal variation between two consecutive BEV feature maps. By analyzing the differences and changes in the BEV feature maps over time, MFANet captures motion information and integrates it into the current feature representation, enabling more robust and accurate detection of 3D objects. Experimental evaluations on the nuScenes benchmark dataset demonstrate that the proposed TM3DOD method achieved significant improvements in 3D detection performance compared with the baseline methods. Additionally, our method achieved comparable performance to state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

In autonomous driving, the fusion of multiple sensors is considered essential to improve the accuracy and safety of 3D object detection. Currently, a fusion scheme combining low-cost cameras with highly robust radars can counteract the performance degradation caused by harsh environments. In this paper, we propose the IRBEVF-Q model, which mainly consists of BEV (Bird's Eye View) fusion coding module and an object decoder module.The BEV fusion coding module solves the problem of unified representation of different modal information by fusing the image and radar features through 3D spatial reference points as a medium. The query in the object decoder, as a core component, plays an important role in detection. In this paper, Heat Map-Guided Query Initialization (HGQI) and Dynamic Position Encoding (DPE) are proposed in query construction to increase the a priori information of the query. The Auxiliary Noise Query (ANQ) then helps to stabilize the matching. The experimental results demonstrate that the proposed fusion model IRBEVF-Q achieves an NDS of 0.575 and a mAP of 0.476 on the nuScenes test set. Compared to recent state-of-the-art methods, our model shows significant advantages, thus indicating that our approach contributes to improving detection accuracy. [ABSTRACT FROM AUTHOR]

The bird's-eye view (BEV) method, which is a vision-centric representation-based perception task, is essential and promising for future Autonomous Vehicle perception. It has advantages of fusion-friendly, intuitive, end-to-end optimization and is cheaper than LiDAR. The performance of existing BEV methods, however, would be deteriorated under the situation of a tire blow-out. This is because they quite rely on accurate camera calibration which may be disabled by noisy camera parameters during blow-out. Therefore, it is extremely unsafe to use existing BEV methods in the tire blow-out situation. In this paper, we propose a geometry-guided auto-resizable kernel transformer (GARKT) method, which is designed especially for vehicles with tire blow-out. Specifically, we establish a camera deviation model for vehicles with tire blow-out. Then we use the geometric priors to attain the prior position in perspective view with auto-resizable kernels. The resizable perception areas are encoded and flattened to generate BEV representation. GARKT predicts the nuScenes detection score (NDS) with a value of 0.439 on a newly created blow-out dataset based on nuScenes. NDS can still obtain 0.431 when the tire is completely flat, which is much more robust compared to other transformer-based BEV methods. Moreover, the GARKT method has almost real-time computing speed, with about 20.5 fps on one GPU. [ABSTRACT FROM AUTHOR]

Accurate and comprehensive 3D objects detection is important for perception systems in autonomous driving. Nevertheless, contemporary mainstream methods tend to perform more effectively on large objects in regions proximate to the LiDAR, leaving limited exploration of long-range objects and small objects. The divergent point pattern of LiDAR, which results in a reduction in point density as the distance increases, leads to a non-uniform point distribution that is ill-suited to discretized volumetric feature extraction. To address this challenge, we propose the Foreground Voxel Proposal (FVP) module, which effectively locates and generates voxels at the foreground of objects. The outputs are subsequently merged to mitigating the difference in point cloud density and completing the object shape. Furthermore, the susceptibility of small objects to occlusion results in the loss of feature space. To overcome this, we propose the Multi-Scale Feature Integration Network (MsFIN), which captures contextual information at different ranges. Subsequently, the outputs of these features are integrated through a cascade framework based on transformers in order to supplement the object features space. The extensive experimental results demonstrate that our network achieves remarkable results. Remarkably, our approach demonstrated an improvement of 8.56% AP on the SECOND baseline for the Car detection task at a distance of more than 20 m, and 9.38% AP on the Cyclist detection task. [ABSTRACT FROM AUTHOR]

In complex traffic environments, 3D target tracking and detection are often occluded by various stationary and moving objects. When the target is occluded, its apparent characteristics change, resulting in a decrease in the accuracy of tracking and detection. In order to solve this problem, we propose to learn the vehicle behavior from the driving data, predict and calibrate the vehicle trajectory, and finally use the artificial fish swarm algorithm to optimize the tracking results. The experiments show that compared with the CenterTrack method, the proposed method improves the key indicators of MOTA (Multi-Object Tracking Accuracy) in 3D object detection and tracking on the nuScenes dataset, and the frame rate is 26 fps. [ABSTRACT FROM AUTHOR]

Autonomous vehicles (AVs) strive to adapt to the specific characteristics of sustainable urban environments. Accurate 3D object detection with LiDAR is paramount for autonomous driving. However, existing research predominantly relies on the 3D object-based assumption, which overlooks the complexity of real-world road environments. Consequently, current methods experience performance degradation when targeting only local features and overlooking the intersection of objects and road features, especially in uneven road conditions. This study proposes a 3D Object-Oriented-Segmentation Spatial-Attention ( O 2 SAT) approach to distinguish object points from road points and enhance the keypoint feature learning by a channel-wise spatial attention mechanism. O 2 SAT consists of three modules: Object-Oriented Segmentation (OOS), Spatial-Attention Feature Reweighting (SFR), and Road-Aware 3D Detection Head (R3D). OOS distinguishes object and road points and performs object-aware downsampling to augment data by learning to identify the hidden connection between landscape and object; SFR performs weight augmentation to learn crucial neighboring relationships and dynamically adjust feature weights through spatial attention mechanisms, which enhances the long-range interactions and contextual feature discrimination for noise suppression, improving overall detection performance; and R3D utilizes refined object segmentation and optimized feature representations. Our system forecasts prediction confidence into existing point-backbones. Our method's effectiveness and robustness across diverse datasets (KITTI) has been demonstrated through vast experiments. The proposed modules seamlessly integrate into existing point-based frameworks, following a plug-and-play approach. [ABSTRACT FROM AUTHOR]

The accuracy and real-time performance of 3D object detection are key factors limiting its widespread application. While cameras capture detailed color and texture features, they lack depth information compared to LiDAR. Multimodal detection combining both can improve results but incurs significant computational overhead, affecting real-time performance. To address these challenges, this paper presents a real-time multimodal fusion model called Fast Transfusion that combines the benefits of LiDAR and camera sensors and reduces the computational burden of their fusion. Specifically, our Fast Transfusion method uses QConv (Quick Convolution) to replace the convolutional backbones compared to other models. QConv concentrates the convolution operations at the feature map center, where the most information resides, to expedite inference. It also utilizes deformable convolution to better match the actual shapes of detected objects, enhancing accuracy. And the model incorporates EH Decoder (Efficient and Hybrid Decoder) which decouples multiscale fusion into intra-scale interaction and cross-scale fusion, efficiently decoding and integrating features extracted from multimodal data. Furthermore, our proposed semi-dynamic query selection refines the initialization of object queries. On the KITTI 3D object detection dataset, our proposed approach reduced the inference time by 36 ms and improved 3D AP by 1.81% compared to state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Recently Object detection and tracking using fusion of LiDAR and RGB camera for the autonomous vehicle environment is a challenging task. The existing works initiates several object detection and tracking frameworks using Artificial Intelligence (AI) algorithms. However, they were limited with high false positives and computation time issues thus lacking the performance of autonomous driving environment. The existing issues are resolved by proposing Hybrid Deep Learning based Multi Object Detection and Tracking (HDL-MODT) using sensor fusion methods. The proposed work performs fusion of solid state LiDAR, Pseudo LiDAR, and RGB camera for improving detection and tracking quality. At first, the multi-stage preprocessing is done in which noise removal is performed using Adaptive Fuzzy Filter (A-Fuzzy). The pre-processed fused image is then provided for instance segmentation to reduce the classification and tracking complexity. For that, the proposed work adopts Lightweight General Adversarial Networks (LGAN). The segmented image is provided for object detection and tracking using HDL. For reducing the complexity, the proposed work utilized VGG-16 for feature extraction which forms the feature vectors. The features vectors are then provided for object detection using YOLOv4. Finally, the detected objects were tracked using Improved Unscented Kalman Filter (IUKF) and mapping the vehicles using time based mapping by considering their RFID, velocity, location, dimension and unique ID. The simulation of the proposed work is carried out using MATLAB R2020a simulation tool and performance of the proposed work is compared with several metrics that show that the proposed work outperforms than the existing works. [ABSTRACT FROM AUTHOR]

Point clouds and RGB images are both critical data for 3D object detection. While recent multi-modal methods combine them directly and show remarkable performances, they ignore the distinct forms of these two types of data. For mitigating the influence of this intrinsic difference on performance, we propose a novel but effective fusion model named LI-Attention model, which takes both RGB features and point cloud features into consideration and assigns a weight to each RGB feature by attention mechanism. Furthermore, based on the LI-Attention model, we propose a 3D object detection method called image attention transformer network (IAT-Net) specialized for indoor RGB-D scene. Compared with previous work on multi-modal detection, IAT-Net fuses elaborate RGB features from 2D detection results with point cloud features in attention mechanism, meanwhile generates and refines 3D detection results with transformer model. Extensive experiments demonstrate that our approach outperforms state-of-the-art performance on two widely used benchmarks of indoor 3D object detection, SUN RGB-D and NYU Depth V2, while ablation studies have been provided to analyze the effect of each module. And the source code for the proposed IAT-Net is publicly available at https://github.com/wisper181/IAT-Net. [ABSTRACT FROM AUTHOR]

The rapid growth on the amount of generated 3D data, particularly in the form of Light Detection And Ranging (LiDAR) point clouds (PCs), poses very significant challenges in terms of data storage, transmission, and processing. Point cloud (PC) representation of 3D visual information has shown to be a very flexible format with many applications ranging from multimedia immersive communication to machine vision tasks in the robotics and autonomous driving domains. In this paper, we investigate the performance of four reference 3D object detection techniques, when the input PCs are compressed with varying levels of degradation. Compression is performed using two MPEG standard coders based on 2D projections and octree decomposition, as well as two coding methods based on Deep Learning (DL). For the DL coding methods, we used a Joint Photographic Experts Group (JPEG) reference PC coder, that we adapted to accept LiDAR PCs in both Cartesian and cylindrical coordinate systems. The detection performance of the four reference 3D object detection methods was evaluated using both pre-trained models and models specifically trained using degraded PCs reconstructed from compressed representations. It is shown that LiDAR PCs can be compressed down to 6 bits per point with no significant degradation on the object detection precision. Furthermore, employing specifically trained detection models improves the detection capabilities even at compression rates as low as 2 bits per point. These results show that LiDAR PCs can be coded to enable efficient storage and transmission, without significant object detection performance loss. [ABSTRACT FROM AUTHOR]

Recently, multi-modal 3D object detection techniques based on point clouds and images have received increasing attention. However, existing methods for multi-modal feature fusion are often relatively singular, and single point cloud representation methods also have some limitations. For example, voxelization may result in the loss of fine-grained information, while 2D images lack depth information, which can restrict the accuracy of detection. Therefore, in this work, we propose a novel method for multi-modal 3D object detection based on point cloud diversity representation and dual feature fusion, PCDR-DFF, to improve the prediction accuracy of 3D object detection. Firstly, point clouds are projected to the image coordinate system and extract multi-level features of the point cloud corresponding to the image using a 2D backbone network. Then, the point clouds are jointly characterized using graphs and pillars, and the 3D features of the point clouds are extracted using graph neural networks and residual connectivity. Finally, a dual feature fusion method is designed to improve the accuracy of detection with the help of a well-designed multi-point fusion model and multi-feature fusion mechanism embedded with a spare 3D-U Net. Extensive experiments on the KITTI dataset demonstrate the effectiveness and competitiveness of our proposed models in comparison with other methods. [ABSTRACT FROM AUTHOR]

Aiming at the problem of low detection accuracy caused by insufficient target feature extraction in substation 3D scene, a voxelization-based 3D object detection model AugSecond for substation scene is proposed, which is designed based on the Second network structure. It introduces a triple attention mechanism in the voxel feature encoding stage, which focuses on multi-dimensional attention to enhance the key information of the target and reduce the interference of irrelevant feature information. It designes asymmetric sparse convolutional networks, uses asymmetric convolution to improve convolutional kernel representation capabilities and fuses multi-scale features to enrich target geometry information. Meanwhile, the position regression loss is optimized, and CIoU Loss is used to further consider the geometric correlation between bounding boxes to speed up the network convergence. Experiments on self-built power scene data sets and public data sets show that compared with the benchmark model, AugSecond model significantly improves recognition accuracy and has real-time reasoning speed, which proves the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

To address the problem that image and point cloud features are fused in a coarse fusion way and cannot achieve deep fusion, this paper proposes a multimodal 3D object detection architecture based on a mutual feature gating mechanism. First, since the feature aggregation approach based on the set abstraction layer cannot obtain fine-grained features, a point-based self-attention mechanism module is designed. This module is added to the extraction branch of point cloud features to achieve fine-grained feature aggregation while maintaining accurate location information. Second, a new gating mechanism is designed for the deep fusion of image and point cloud. Deep fusion is achieved by mutual feature weighting between the image and the point cloud. The newly fused features are then fed into a feature refinement network to extract classification confidence and 3D target bounding boxes. Finally, a multi-scale detection architecture is proposed to obtain a more complete object shape. The location-based encoding feature algorithm is also designed to focus the interest points in the region of interest adaptively. The whole architecture shows outstanding performance on the KITTI3D and nuSenece datasets, especially at the difficult level. It shows that the framework solves the problem of low detection rates in LiDAR mode due to the low number of surface points obtained from distant objects. [ABSTRACT FROM AUTHOR]

LiDAR-based 3D object detection and localization are crucial components of autonomous navigation systems, including autonomous vehicles and mobile robots. Most existing LiDAR-based 3D object detection and localization approaches primarily use geometric or structural feature abstractions from LiDAR point clouds. However, these approaches can be susceptible to environmental noise due to adverse weather conditions or the presence of highly scattering media. In this work, we propose an intensity-aware voxel encoder for robust 3D object detection. The proposed voxel encoder generates an intensity histogram that describes the distribution of point intensities within a voxel and is used to enhance the voxel feature set. We integrate this intensity-aware encoder into an efficient single-stage voxel-based detector for 3D object detection. Experimental results obtained using the KITTI dataset show that our method achieves comparable results with respect to the state-of-the-art method for car objects in 3D detection and from a bird's-eye view and superior results for pedestrian and cyclic objects. Furthermore, our model can achieve a detection rate of 40.7 FPS during inference time, which is higher than that of the state-of-the-art methods and incurs a lower computational cost. [ABSTRACT FROM AUTHOR]

In the realm of computer vision, the integration of advanced techniques into the pre-processing of RGB-D camera inputs poses a significant challenge, given the inherent complexities arising from diverse environmental conditions and varying object appearances. Therefore, this paper introduces FusionVision, an exhaustive pipeline adapted for the robust 3D segmentation of objects in RGB-D imagery. Traditional computer vision systems face limitations in simultaneously capturing precise object boundaries and achieving high-precision object detection on depth maps, as they are mainly proposed for RGB cameras. To address this challenge, FusionVision adopts an integrated approach by merging state-of-the-art object detection techniques, with advanced instance segmentation methods. The integration of these components enables a holistic (unified analysis of information obtained from both color RGB and depth D channels) interpretation of RGB-D data, facilitating the extraction of comprehensive and accurate object information in order to improve post-processes such as object 6D pose estimation, Simultanious Localization and Mapping (SLAM) operations, accurate 3D dataset extraction, etc. The proposed FusionVision pipeline employs YOLO for identifying objects within the RGB image domain. Subsequently, FastSAM, an innovative semantic segmentation model, is applied to delineate object boundaries, yielding refined segmentation masks. The synergy between these components and their integration into 3D scene understanding ensures a cohesive fusion of object detection and segmentation, enhancing overall precision in 3D object segmentation. [ABSTRACT FROM AUTHOR]

Facing the significant challenge of 3D object detection in complex weather conditions and road environments, existing algorithms based on single-frame point cloud data struggle to achieve desirable results. These methods typically focus on spatial relationships within a single frame, overlooking the semantic correlations and spatiotemporal continuity between consecutive frames. This leads to discontinuities and abrupt changes in the detection outcomes. To address this issue, this paper proposes a multi-frame 3D object detection algorithm based on a deformable spatiotemporal Transformer. Specifically, a deformable cross-scale Transformer module is devised, incorporating a multi-scale offset mechanism that non-uniformly samples features at different scales, enhancing the spatial information aggregation capability of the output features. Simultaneously, to address the issue of feature misalignment during multi-frame feature fusion, a deformable cross-frame Transformer module is proposed. This module incorporates independently learnable offset parameters for different frame features, enabling the model to adaptively correlate dynamic features across multiple frames and improve the temporal information utilization of the model. A proposal-aware sampling algorithm is introduced to significantly increase the foreground point recall, further optimizing the efficiency of feature extraction. The obtained multi-scale and multi-frame voxel features are subjected to an adaptive fusion weight extraction module, referred to as the proposed mixed voxel set extraction module. This module allows the model to adaptively obtain mixed features containing both spatial and temporal information. The effectiveness of the proposed algorithm is validated on the KITTI, nuScenes, and self-collected urban datasets. The proposed algorithm achieves an average precision improvement of 2.1% over the latest multi-frame-based algorithms. [ABSTRACT FROM AUTHOR]

The preprocessing of point cloud data has always been an important problem in 3D object detection. Due to the large volume of point cloud data, voxelization methods are often used to represent the point cloud while reducing data density. However, common voxelization randomly selects sampling points from voxels, which often fails to represent local spatial features well due to noise. To preserve local features, this paper proposes an optimized voxel downsampling(OVD) method based on evidence theory. This method uses fuzzy sets to model basic probability assignments (BPAs) for each candidate point, incorporating point location information. It then employs evidence theory to fuse the BPAs and determine the selected sampling points. In the PointPillars 3D object detection algorithm, the point cloud is partitioned into pillars and encoded using each pillar’s points. Convolutional neural networks are used for feature extraction and detection. Another contribution is the proposed improved PointPillars based on evidence theory (ET-PointPillars) by introducing an OVD-based feature point sampling module in the PointPillars’ pillar feature network, which can select feature points in pillars using the optimized method, computes offsets to these points, and adds them as features to facilitate learning more object characteristics, improving traditional PointPillars. Experiments on the KITTI datasets validate the method’s ability to preserve local spatial features. Results showed improved detection precision, with a 2.73 % average increase for pedestrians and cyclists on KITTI. [ABSTRACT FROM AUTHOR]

Lidar and camera are essential sensors for environment perception in autonomous driving. However, fully fusing heterogeneous data from multiple sources remains a non-trivial challenge. As a result, 3D object detection based on multi-modal sensor fusion are often inferior to single-modal methods only based on Lidar, which indicates that multi-sensor machine vision still needs development. In this paper, we propose an adaptive fusion module based on cross-modal transformer block(AFMCT) for 3D object detection by utilizing a bidirectional enhancing strategy. Specifically, we first enhance image feature by extracting an attention-based point feature based on a cross-modal transformer block and linking them in a concatenation fashion, followed by another cross-modal transformer block acting on the enhanced image feature to strengthen the point feature with image semantic information. Extensive experiments operated on the 3D detection benchmark of the KITTI dataset reveal that our proposed structure can significantly improve the detection accuracy of Lidar-only methods and outperform the existing advanced multi-sensor fusion modules by at least 0.45%, which indicates that our method might be a feasible solution to improving 3D object detection based on multi-sensor fusion. [ABSTRACT FROM AUTHOR]

Recent years have witnessed significant advancements in deep learning-based 3D object detection, leading to its widespread adoption in numerous applications. As 3D object detectors become increasingly crucial for security-critical tasks, it is imperative to understand their robustness against adversarial attacks. This paper presents the first comprehensive evaluation and analysis of the robustness of LiDAR-based 3D detectors under adversarial attacks. Specifically, we extend three distinct adversarial attacks to the 3D object detection task, benchmarking the robustness of state-of-the-art LiDAR-based 3D object detectors against attacks on the KITTI and Waymo datasets. We further analyze the relationship between robustness and detector properties. Additionally, we explore the transferability of cross-model, cross-task, and cross-data attacks. Thorough experiments on defensive strategies for 3D detectors are conducted, demonstrating that simple transformations like flipping provide little help in improving robustness when the applied transformation strategy is exposed to attackers. Finally, we propose balanced adversarial focal training, based on conventional adversarial training, to strike a balance between accuracy and robustness. Our findings will facilitate investigations into understanding and defending against adversarial attacks on LiDAR-based 3D object detectors, thus advancing the field. The source code is publicly available at https://github.com/Eaphan/Robust3DOD. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer-Aided Design & Computer Graphics / Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao is the property of Gai Kan Bian Wei Hui and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer-Aided Design & Computer Graphics / Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao is the property of Gai Kan Bian Wei Hui and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)