CornerPoint3D: Look at the Nearest Corner Instead of the Center
- URL: http://arxiv.org/abs/2504.02464v1
- Date: Thu, 03 Apr 2025 10:33:43 GMT
- Title: CornerPoint3D: Look at the Nearest Corner Instead of the Center
- Authors: Ruixiao Zhang, Runwei Guan, Xiangyu Chen, Adam Prugel-Bennett, Xiaohao Cai,
- Abstract summary: 3D object detection aims to predict object centers, dimensions, and rotations from LiDAR point clouds.<n>LiDAR captures only the near side of objects, making center-based detectors prone to poor localization accuracy in cross-domain tasks.<n>We propose a novel 3D object detector, coined as CornerPoint3D, which is built upon CenterPoint and uses heatmaps to supervise the learning and detection of the nearest corner of each object.
- Score: 7.293031759018836
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: 3D object detection aims to predict object centers, dimensions, and rotations from LiDAR point clouds. Despite its simplicity, LiDAR captures only the near side of objects, making center-based detectors prone to poor localization accuracy in cross-domain tasks with varying point distributions. Meanwhile, existing evaluation metrics designed for single-domain assessment also suffer from overfitting due to dataset-specific size variations. A key question arises: Do we really need models to maintain excellent performance in the entire 3D bounding boxes after being applied across domains? Actually, one of our main focuses is on preventing collisions between vehicles and other obstacles, especially in cross-domain scenarios where correctly predicting the sizes is much more difficult. To address these issues, we rethink cross-domain 3D object detection from a practical perspective. We propose two new metrics that evaluate a model's ability to detect objects' closer-surfaces to the LiDAR sensor. Additionally, we introduce EdgeHead, a refinement head that guides models to focus more on learnable closer surfaces, significantly improving cross-domain performance under both our new and traditional BEV/3D metrics. Furthermore, we argue that predicting the nearest corner rather than the object center enhances robustness. We propose a novel 3D object detector, coined as CornerPoint3D, which is built upon CenterPoint and uses heatmaps to supervise the learning and detection of the nearest corner of each object. Our proposed methods realize a balanced trade-off between the detection quality of entire bounding boxes and the locating accuracy of closer surfaces to the LiDAR sensor, outperforming the traditional center-based detector CenterPoint in multiple cross-domain tasks and providing a more practically reasonable and robust cross-domain 3D object detection solution.
Related papers
- Detect Closer Surfaces that can be Seen: New Modeling and Evaluation in Cross-domain 3D Object Detection [7.464834150824093]
We propose two metrics to measure 3D object detection models' ability of detecting the closer surfaces to the sensor on the ego vehicle.
We also propose a refinement head, named EdgeHead, to guide models to focus more on the learnable closer surfaces.
arXiv Detail & Related papers (2024-07-04T17:06:16Z) - DiffuBox: Refining 3D Object Detection with Point Diffusion [74.01759893280774]
We introduce a novel diffusion-based box refinement approach to ensure robust 3D object detection and localization.<n>This method employs a domain-agnostic diffusion model conditioned on the LiDAR points surrounding a coarse bounding box, to simultaneously refine the box's location, size, and orientation.<n>Our results reveal significant improvements across different datasets, object classes and detectors.
arXiv Detail & Related papers (2024-05-25T03:14:55Z) - An Empirical Analysis of Range for 3D Object Detection [70.54345282696138]
We present an empirical analysis of far-field 3D detection using the long-range detection dataset Argoverse 2.0.
Near-field LiDAR measurements are dense and optimally encoded by small voxels, while far-field measurements are sparse and are better encoded with large voxels.
We propose simple techniques to efficiently ensemble models for long-range detection that improve efficiency by 33% and boost accuracy by 3.2% CDS.
arXiv Detail & Related papers (2023-08-08T05:29:26Z) - AGO-Net: Association-Guided 3D Point Cloud Object Detection Network [86.10213302724085]
We propose a novel 3D detection framework that associates intact features for objects via domain adaptation.
We achieve new state-of-the-art performance on the KITTI 3D detection benchmark in both accuracy and speed.
arXiv Detail & Related papers (2022-08-24T16:54:38Z) - 3D Object Detection Combining Semantic and Geometric Features from Point
Clouds [19.127930862527666]
We propose a novel end-to-end two-stage 3D object detector named SGNet for point clouds scenes.
The VTPM is a Voxel-Point-Based Module that finally implements 3D object detection in point space.
As of September 19, 2021, for KITTI dataset, SGNet ranked 1st in 3D and BEV detection on cyclists with easy difficulty level, and 2nd in the 3D detection of moderate cyclists.
arXiv Detail & Related papers (2021-10-10T04:43:27Z) - Anchor-free 3D Single Stage Detector with Mask-Guided Attention for
Point Cloud [79.39041453836793]
We develop a novel single-stage 3D detector for point clouds in an anchor-free manner.
We overcome this by converting the voxel-based sparse 3D feature volumes into the sparse 2D feature maps.
We propose an IoU-based detection confidence re-calibration scheme to improve the correlation between the detection confidence score and the accuracy of the bounding box regression.
arXiv Detail & Related papers (2021-08-08T13:42:13Z) - Delving into Localization Errors for Monocular 3D Object Detection [85.77319416168362]
Estimating 3D bounding boxes from monocular images is an essential component in autonomous driving.
In this work, we quantify the impact introduced by each sub-task and find the localization error' is the vital factor in restricting monocular 3D detection.
arXiv Detail & Related papers (2021-03-30T10:38:01Z) - M3DSSD: Monocular 3D Single Stage Object Detector [82.25793227026443]
We propose a Monocular 3D Single Stage object Detector (M3DSSD) with feature alignment and asymmetric non-local attention.
The proposed M3DSSD achieves significantly better performance than the monocular 3D object detection methods on the KITTI dataset.
arXiv Detail & Related papers (2021-03-24T13:09:11Z) - 1st Place Solution for Waymo Open Dataset Challenge -- 3D Detection and
Domain Adaptation [7.807118356899879]
We propose a one-stage, anchor-free and NMS-free 3D point cloud object detector AFDet.
AFDet serves as a strong baseline in our winning solution.
We design stronger networks and enhance the point cloud data using densification and point painting.
arXiv Detail & Related papers (2020-06-28T04:49:39Z) - Center-based 3D Object Detection and Tracking [8.72305226979945]
Three-dimensional objects are commonly represented as 3D boxes in a point-cloud.
This representation mimics the well-studied image-based 2D bounding-box detection but comes with additional challenges.
In this paper, we propose to represent, detect, and track 3D objects as points.
Our framework, CenterPoint, first detects centers of objects using a keypoint detector and regresses to other attributes, including 3D size, 3D orientation, and velocity.
The resulting detection and tracking algorithm is simple, efficient, and effective.
arXiv Detail & Related papers (2020-06-19T17:59:39Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.