DZDiankun Zhang

Diankun Zhang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Diankun Zhang has a total of 34 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Top Cited Papers

    SparseAD: Sparse Query-Centric Paradigm for Efficient End-to-End Autonomous Driving

    by Runwen Zhu, Jianbo Zhao, Diankun Zhang, Guoan Wang, Xiwu Chen, Siyu Zhang, Jiahao Gong, Qibin Zhou, Wenyuan Zhang, Ningzi Wang, et al.

    End-to-End paradigms use a unified framework to implement multi-tasks in an autonomous driving system. Despite simplicity and clarity, the performance of end-to-end autonomous driving methods on sub-tasks is still far behind the single-task methods. Meanwhile, the widely used dense BEV features in previous end-to-end methods make it costly to extend to more modalities or tasks. In this paper, we propose a Sparse query-centric paradigm for end-to-end Autonomous Driving (SparseAD), where the sparse queries completely represent the whole driving scenario across space, time and tasks without any dense BEV representation. Concretely, we design a unified sparse architecture for perception tasks including detection, tracking, and online mapping. Moreover, we revisit motion prediction and planning, and devise a more justifiable motion planner framework. On the challenging nuScenes dataset, SparseAD achieves SOTA full-task performance among end-to-end methods and significantly narrows the performance gap between end-to-end paradigms and single-task methods.


  • Open Access

    PillarNeSt: Embracing Backbone Scaling and Pretraining for Pillar-Based 3D Object Detection

    by Weixin Mao, Tiancai Wang, Diankun Zhang, Junjie Yan, Osamu Yoshie

    Pillar-based 3D object detectors mainly employ randomly initialized 2D convolution neural network (ConvNet) for feature extraction and fail to enjoy the benefits from the backbone scaling and pretraining in the image domain. This paper shows the effectiveness of 2D backbone scaling and pretraining for pillar-based 3D object detectors. For better backbone scaling, we first introduce several design principles for point cloud backbone, to tackle the sparsity of point cloud and improve the effective receptive field. The backbone scaling is achieved by adaptively designed based on the model size. For backbone pretraining, we propose a weight adaptation module, to transfer the image knowledge obtained by pretraining on large-scale image datasets for the point cloud. Our proposed pillar-based detector, termed PillarNeSt, outperforms the existing 3D object detectors by a large margin on the nuScenes and Argoversev2 datasets. Code is released at https://github.com/WayneMao/PillarNeSt .


  • Fully Sparse Transformer 3-D Detector for LiDAR Point Cloud

    by Diankun Zhang, Zhijie Zheng, Haoyu Niu, Xueqing Wang, Xiaojun Liu

    The 3D object detector usually uses a framework similar to 2D detection and benefits from the advancements of 2D detection tasks. In these frameworks, it is necessary to make the unstructured, sparse point cloud features into dense grids to be compatible with popular 2D operators such as convolution and transformers, which also causes extra computational costs. In this paper, we propose a simple and efficient Fully Sparse TRansformer (FSTR) for LiDAR-based 3D object detection, which is able to combine with state-of-the-art sparse backbones to form a fully sparse, end-to-end, simple, and efficient detection framework. FSTR uses the sparse voxel feature from the sparse backbone as the input token without any custom operators. Further, we introduce the dynamic queries to provide a priori location and context of the foreground for the decoder and drop the high-confidence background tokens to further reduce redundant computations. We propose Gaussian denoising queries to speed up the decoder training and make it more adaptable to the distribution of sparse voxel features. Extensive experiments on the nuScenes benchmark and the Argoverse2 benchmark validate the effectiveness of the proposed method. FSTR outperforms all LiDAR real-time methods by 69.5 mAP and 72.9 NDS on the official benchmark of nuScenes dataset. On the long-range detection benchmark Argoverse2, the proposed method achieves a new state-of-art performance of 39.9 mAP which outperforms the existing LiDAR detectors, even the LiDAR-Camera detectors by a large margin (+9.4 mAP and +7.5mAP), showing the great advantage of the proposed method for long-range detection.


  • Open Access

    Segmentation Can Aid Detection: Segmentation-Guided Single Stage Detection for 3D Point Cloud

    by Xueqing Wang, Diankun Zhang, Haoyu Niu, Xiaojun Liu

    Detecting accurate 3D bounding boxes from point cloud data plays an essential role in autonomous driving. However, improving performance requires more complex models, which come with high memory and computational cost. In this work, we design a Segmentation-Guided Auxiliary Network (SGAN) to improve the localization quality of detection. The points from different levels are concatenated to generate the multi-scale feature for the points used for prediction, i.e., candidate points. SGAN is jointly optimized by two tasks of candidate points—segmentation and center estimation—and it is only used in training and therefore introduces no extra computation in the inference stage. Furthermore, we consider that point-based detectors suffer from the outline points of sampling, so we explore the correlation between the data and propose the Point Cloud External Attention (PCEA) to extract the semantic features with a low memory cost. Our method SGSSD achieves a large margin against the baseline on the KITTI and Waymo datasets, and it runs at 25 FPS for inference on the KITTI test set with a single NVIDIA RTX 3090.


  • ARFA: Adaptive Reception Field Aggregation for 3-D Detection From LiDAR Point Cloud

    by Diankun Zhang, Xueqing Wang, Zhijie Zheng, Xiaojun Liu, Guangyou Fang

    Submanifold convolution is widely used in 3D detection. However, it brings different receptive fields to voxels due to the non-uniform distribution in LiDAR point clouds, resulting in degradation of the feature extraction ability for distant voxels and the performance of detectors. We propose a solution, Adaptive Receptive Field Aggregation (ARFA) network, an end-to-end two-stage LiDAR 3D object detection architecture. ARFA searches the top-k nearest neighbors to adaptively adjust the receptive field of sparse voxels, followed by a self-attention aggregation module with density feature embedding to aggregate the semantic information in the receptive field. In order to further strengthen the detection performance for small objects, we also propose an Up-Sampling Bird’s Eyes View (U-BEV) Backbone and an IoU-Aware Head to enhance the quality of the proposals and rectify the confidence of the predicted bounding boxes. ARFA outperforms the state-of-the-art methods on the Waymo Open Dataset and achieves competitive results on the popular KITTI Dataset.


  • Unsupervised Domain Adaptive 3-D Detection With Data Adaption From LiDAR Point Cloud

    by Diankun Zhang, Xueqing Wang, Zhijie Zheng, Xiaojun Liu

    Existing unsupervised domain adaptive (UDA) 3D detection methods only address the domain gap caused by the prior size of 3D bounding boxes between different datasets, which ignore the difference in the distribution of point clouds. To address this challenge, we propose an unsupervised domain adaptive 3D detection by data adaption, which trains the model by transferring the source domain instances into the target domain scenes by adaptive point distribution. First, an instance transferring method is proposed for selecting and transferring suitable instances from the source domain into the target domain scene; Second, we propose an adaptive downsampling method to adjust the point cloud distribution of the transferred instances to approximate the points distribution of the target domain. Finally, our method trains the randomly initialized detector with the pseudo-instances in the target domain. To the best of our knowledge, we first address the UDA problem of the 3D detectors from the perspective of data. Extensive experiments on several popular datasets show that the proposed method outperforms the existing state-of-the-art methods by a large margin. Further experiments also show our approach is detector-agnostic and achieves consistent and significant gains on all types of 3D detectors.


  • Open Access

    Structure Information is the Key: Self-Attention RoI Feature Extractor in 3D Object Detection

    by Diankun Zhang, Zhijie Zheng, Xueting Bi, Xiaojun Liu

    Unlike 2D object detection where all RoI features come from grid pixels, the RoI feature extraction of 3D point cloud object detection is more diverse. In this paper, we first compare and analyze the differences in structure and performance between the two state-of-the-art models PV-RCNN and Voxel-RCNN. Then, we find that the performance gap between the two models does not come from point information, but structural information. The voxel features contain more structural information because they do quantization instead of downsampling to point cloud so that they can contain basically the complete information of the whole point cloud. The stronger structural information in voxel features makes the detector have higher performance in our experiments even if the voxel features don't have accurate location information. Then, we propose that structural information is the key to 3D object detection. Based on the above conclusion, we propose a Self-Attention RoI Feature Extractor (SARFE) to enhance structural information of the feature extracted from 3D proposals. SARFE is a plug-and-play module that can be easily used on existing 3D detectors. Our SARFE is evaluated on both KITTI dataset and Waymo Open dataset. With the newly introduced SARFE, we improve the performance of the state-of-the-art 3D detectors by a large margin in cyclist on KITTI dataset while keeping real-time capability.

  • Analysis of Aeromagnetic Swing Noise and Corresponding Compensation Method

    by Diankun Zhang, Xiaojun Liu, Xiaodong Qu, Wanhua Zhu, Ling Huang, Guangyou Fang

    Aeromagnetic noise compensation is a vital part of aerial survey measurement, and its compensation effect directly determines the quality of aeromagnetic survey data. At present, the commonly used compensation model is the T-L model, and the least squares method is used to solve for the coefficients. However, the noise source modeled in the T-L model is incomplete. Since the tail boom cannot be completely rigid, tail-boom swing is an unavoidable problem in aeromagnetic measurement. This kind of swing is the most obvious when the aircraft is maneuvering, and it will significantly interfere with the measurement data of the sensor. In this article, two causes of the swing noise are analyzed, and the nonlinear relationship between the swing displacement and the noise is derived. Since it is difficult to express the nonlinear relationship with mathematical forms to compensate for the aeromagnetic data, we propose a new compensation method that uses a 1-D convolutional neural network to perform secondary compensation on the data already compensated by the T-L model in order to remove the effect of tail-boom swing. The flight experiment data show that the proposed method can significantly improve the quality of aeromagnetic data. Compared with the T-L method, the improve ratio is increased by 60%–100%. It shows that the proposed method has a remarkable compensation effect for aeromagnetic noise.