RangeDet:In Defense of Range View for LiDAR-based 3D Object Detection

Fan, Lue; Xiong, Xuan; Wang, Feng; Wang, Naiyan; Zhang, Zhaoxiang

doi:10.48550/arxiv.2103.10039

Cited by 16 publications

(25 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Without bells and whistles, our approach works better than PV-RCNN [39]. Furthermore, we compare our framework on the vehicle class for different distances with state-of-the-art methods, including StarNet [30], PointPillars [19], RCD [1], Det3D [63], RangeDet [10] and PV-RCNN [39]. As shown in Table 2, M3DETR outperforms PV-RCNN significantly in both LEVEL 1 and LEVEL 2 difficulty levels across all distances, demonstrating the effectiveness of the newly proposed framework.…”

Section: Resultsmentioning

confidence: 99%

M3DeTR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers

Guan¹,

Wang²,

Lan³

et al. 2021

Preprint

View full text Add to dashboard Cite

We present a novel architecture for 3D object detection, M3DETR, which combines different point cloud representations (raw, voxels, bird-eye view) with different feature scales based on multi-scale feature pyramids. M3DETR is the first approach that unifies multiple point cloud representations, feature scales, as well as models mutual relationships between point clouds simultaneously using transformers. We perform extensive ablation experiments that highlight the benefits of fusing representation and scale, and modeling the relationships. Our method achieves state-of-the-art performance on the KITTI 3D object detection dataset and Waymo Open Dataset. Results show that M3DETR improves the baseline significantly by 1.48% mAP for all classes on Waymo Open Dataset. In particular, our approach ranks 1 st on the well-known KITTI 3D Detection Benchmark for both car and cyclist classes, and ranks 1 st on Waymo Open Dataset with single frame point cloud input.

show abstract

Section: Resultsmentioning

confidence: 99%

M3DeTR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers

Guan¹,

Wang²,

Lan³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…PIXOR (Yang, Luo, and Urtasun 2018) assigns all pixels inside ground truth bounding boxes as positive samples in the BEV feature map. Similarly, some range-view-based methods assign pixels on range view maps that are inside 3D ground truth boxes as positive samples (Fan et al 2021;Meyer et al 2019;Bewley et al 2020). Multiple anchor-free detection networks formulate the detection problem as a keypoint detection problem (Ge et al 2020;Sun et al 2021;Yin, Zhou, and Krahenbuhl 2021b).…”

Section: Anchor-free/anchor-based Lidar Detectormentioning

confidence: 99%

“…Object detection from point clouds has become a practical solution to robotics vision, especially in autonomous driving applications. Like the detection methods on 2D images, the 3D detection methods can also be divided into two groups: single-stage (Ge et al 2020;Zhou and Tuzel 2018;Yan, Mao, and Li 2018;He et al 2020;Zheng et al 2021;Lang et al 2019;Bewley et al 2020;Fan et al 2021;Chen et al 2020a) and two-stage (Shi, Wang, and Li 2019;Yang et al 2019;Qi et al 2017;Shi et al 2020c;Li, Wang, and Wang 2021;Shi et al 2021;Deng et al 2021a;Yin, Zhou, and Krahenbuhl 2021a;Sun et al 2021), in terms of the model structure. The two-stage methods usually show better accuracy (Shi, Wang, and Li 2019;Shi et al 2021;Deng et al 2021a) in the classification confidence and the box regression, than the single-stage methods.…”

Section: Introductionmentioning

confidence: 99%

AFDetV2: Rethinking the Necessity of the Second Stage for Object Detection from Point Clouds

Hu¹,

Ding²,

Ge³

et al. 2021

Preprint

View full text Add to dashboard Cite

There have been two streams in the 3D detection from point clouds: single-stage methods and two-stage methods. While the former is more computationally efficient, the latter usually provides better detection accuracy. By carefully examining the two-stage approaches, we have found that if appropriately designed, the first stage can produce accurate box regression. In this scenario, the second stage mainly rescores the boxes such that the boxes with better localization get selected. From this observation, we have devised a single-stage anchor-free network that can fulfill these requirements. This network, named AFDetV2, extends the previous work by incorporating a self-calibrated convolution block in the backbone, a keypoint auxiliary supervision, and an IoU prediction branch in the multi-task head. We take a simple product of the predicted IoU score with the classification heatmap to form the final classification confidence. The enhanced backbone strengthens the box localization capability, and the rescoring approach effectively joins the object presence confidence and the box regression accuracy. As a result, the detection accuracy is drastically boosted in the single-stage. To evaluate our approach, we have conducted extensive experiments on the Waymo Open Dataset and the nuScenes Dataset. We have observed that our AFDetV2 achieves the state-of-theart results on these two datasets, superior to all the prior arts, including both the single-stage and the two-stage 3D detectors. AFDetV2 won the 1 st place in the Real-Time 3D Detection of the Waymo Open Dataset Challenge 2021. In addition, a variant of our model AFDetV2-Base was entitled the "Most Efficient Model" by the Challenge Sponsor, showing a superior computational efficiency. To demonstrate the generality of this single-stage method, we have also applied it to the first stage of the two-stage networks. Without exception, the results show that with the strengthened backbone and the rescoring approach, the second stage refinement is no longer needed.

show abstract

“…2.1 Non-streaming lidar perception Most lidar perception architectures take inspiration from the image perception literature [23,17,16]. Some single-stage methods typically convert the point cloud into a bird's-eye view image [38,15,31] or a range view image [20,10] and perform detection in those views. The most common paradigm is to convert the lidar point cloud into a BEV image as it offers several advantages like a lack of scale ambiguity, a near lack of occlusion, the ease of fusing HD maps [30] and performing simultaneous detection and trajectory predictions [4,18].…”

Section: Related Workmentioning

confidence: 99%

“…While such self-driving cars typically deploy a wide variety of sensors lidars play a key role due to the accurate range information provided. Driven in part by the availability of benchmark datasets [12,3,26], the last decade has seen tremendous progress in lidar based 3D object detection [38,15,31,20,10]. However, these methods all ignore the fact that most lidar sensors scan the scene sequentially as the lidar rotates around the z-axis.…”

Section: Introductionmentioning

confidence: 99%

PolarStream: Streaming Lidar Object Detection and Segmentation with Polar Pillars

Chen¹,

Vora²,

Beijbom³

2021

Preprint

View full text Add to dashboard Cite

Recent works recognized lidars as an inherently streaming data source and showed that the end-to-end latency of lidar perception models can be reduced significantly by operating on wedge-shaped point cloud sectors rather then the full point cloud. However, due to use of cartesian coordinate systems these methods represent the sectors as rectangular regions, wasting memory and compute. In this work we propose using a polar coordinate system and make two key improvements on this design. First, we increase the spatial context by using multi-scale padding from neighboring sectors: preceding sector from the current scan and/or the following sector from the past scan. Second, we improve the core polar convolutional architecture by introducing feature undistortion and range stratified convolutions. Experimental results on the nuScenes dataset show significant improvements over other streaming based methods. We also achieve comparable results to existing non-streaming methods but with lower latencies.

show abstract

RangeDet:In Defense of Range View for LiDAR-based 3D Object Detection

Cited by 16 publications

References 36 publications

M3DeTR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers

M3DeTR: Multi-representation, Multi-scale, Mutual-relation 3D Object Detection with Transformers

AFDetV2: Rethinking the Necessity of the Second Stage for Object Detection from Point Clouds

PolarStream: Streaming Lidar Object Detection and Segmentation with Polar Pillars

Contact Info

Product

Resources

About