VIN: Voxel-based Implicit Network for Joint 3D Object Detection and Segmentation for Lidars

no code implementations • 7 Jul 2021 • Yuanxin Zhong, Minghan Zhu, Huei Peng

A unified neural network structure is presented for joint 3D object detection and point cloud segmentation in this paper.

3D Object Detection Object +4

Monocular 3D Vehicle Detection Using Uncalibrated Traffic Cameras through Homography

1 code implementation • 29 Mar 2021 • Minghan Zhu, Songan Zhang, Yuanxin Zhong, Pingping Lu, Huei Peng, John Lenneman

This paper proposes a method to extract the position and pose of vehicles in the 3D world from a single traffic camera.

Differentiable Computational Geometry for 2D and 3D machine learning

1 code implementation • 22 Nov 2020 • Yuanxin Zhong

With the growth of machine learning algorithms with geometry primitives, a high-efficiency library with differentiable geometric operators are desired.

BIG-bench Machine Learning

Uncertainty-Aware Voxel based 3D Object Detection and Tracking with von-Mises Loss

1 code implementation • 4 Nov 2020 • Yuanxin Zhong, Minghan Zhu, Huei Peng

Object detection and tracking is a key task in autonomy.

3D Object Detection Object +2

Monocular Depth Prediction through Continuous 3D Loss

2 code implementations • 21 Mar 2020 • Minghan Zhu, Maani Ghaffari, Yuanxin Zhong, Pingping Lu, Zhong Cao, Ryan M. Eustice, Huei Peng

In contrast to the current point-to-point loss evaluation approach, the proposed 3D loss treats point clouds as continuous objects; therefore, it compensates for the lack of dense ground truth depth due to LIDAR's sparsity measurements.

Depth Estimation Depth Prediction

Mcity Data Collection for Automated Vehicles Study

no code implementations • 12 Dec 2019 • Yiqun Dong, Yuanxin Zhong, Wenbo Yu, Minghan Zhu, Pingping Lu, Yeyang Fang, Jiajun Hong, Huei Peng

The main goal of this paper is to introduce the data collection effort at Mcity targeting automated vehicle development.