Fetching the paper…

Reading the bibliography…

2022

MT-Net Submission to the Waymo 3D Detection Leaderboard

Chen, Shaoxiang, Jie, Zequn, Wei, Xiaolin et al.

Original
PDF
Post on X

Understand

In this technical report, we introduce our submission to the Waymo 3D Detection leaderboard.

Our network is based on the Centerpoint architecture, but with significant improvements.
We design a 2D backbone to utilize multi-scale features for better detecting objects with various sizes, together with an optimal transport-based target assignment strategy, which dynamically assigns richer supervision signals to the detection candidates.
We also apply test-time augmentation and model-ensemble for further improvements.

MT-Net Submission to the Waymo 3D Detection Leaderboard · Around

Built on

Submanifold sparse convolutional networks
Original
B. Graham and L. van der Maaten · 2017
Earlier work this paper cites.
Improving convolutional networks with self-calibrated convolutions
J.-J. Liu, Q. Hou, M.-M. Cheng, C. Wang, and J. Feng · 2020
Earlier work this paper cites.
Scalability in perception for autonomous driving: Waymo open dataset
P. Sun, H. Kretzschmar, X. Dotiwalla, A. Chouard, V. Patnaik, P. Tsui, J. Guo, Y. Zhou, Y. Chai, B. Caine, et al · 2020
Earlier work this paper cites.

Similar

Real-time anchor-free single-stage 3d detection with iou-awareness
Original
R. Ge, Z. Ding, Y. Hu, W. Shao, L. Huang, K. Li, and Q. Liu · 2021
Cited alongside, same era.
Ota: Optimal transport assignment for object detection
Z. Ge, S. Liu, Z. Li, O. Yoshie, and J. Sun · 2021
Cited alongside, same era.

Then

Center-based 3d object detection and tracking
T. Yin, X. Zhou, and P. Krahenbuhl · 2021
Later among the works it cites.
Centerpoint++ submission to the waymo real-time 3d detection challenge
T. Yin, X. Zhou, and P. Krahenbuhl · 2021
Later among the works it cites.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…