Search

Your search keyword '"Liu, XiaoJun"' showing total 5,988 results
Start Over Author "Liu, XiaoJun"
5,988 results on '"Liu, XiaoJun"'

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

Author

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

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

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.

Published
2021

99. Laser Driven Fluorescence Emission in a Nitrogen Gas Jet at 100 MHz Repetition Rate

Author

Zhang, Jin, Hua, LinQiang, Yu, ShaoGang, Wang, YanLan, Zhu, MuFeng, Xiao, ZhengRong, Gong, Cheng, and Liu, XiaoJun

Subjects
Physics - Atomic Physics
Abstract

We report the fluorescence emission which is driven by femtosecond laser pulses with a repetition rate of 100 MHz and a center wavelength of 1040 nm in a nitrogen gas jet. The experiment is performed in a femtosecond enhancement cavity coupled with high repetition rate laser for the first time to the best of our knowledge. In contrast to previous observation at low repetition rate with a nitrogen gas jet, where the 391 nm radiation was observed but the 337 nm emission was missing, the 337 nm emission is 3 times stronger than the 391 nm emission in our experiment. By examining the dependence of the radiation intensity on the flow rate of the nitrogen gas and the polarization of the pump pulse, the formation mechanism of the N2(C3{\Pi}u) triplet excited state, i.e., the upper state of the 337 nm emission, is investigated. We attribute the main excitation process to the inelastic collision excitation process, and exclude the possibility of the dissociative recombination as the dominate pathway. The role of the steady state plasma that is generated under our experimental conditions is also discussed., Comment: 7 pages, 6 figures

Published
2021
Full Text
View/download PDF

100. Research on the Practice of VR Entering the Classroom–Taking City University Students of Macau as an Example

Author

Liu, Liang, Liu, Yuzhuo, Lai, Ivan Ka Wai, Liu, Xiaojun, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Li, Chen, editor, Cheung, Simon K. S., editor, Wang, Fu Lee, editor, Lu, Angel, editor, and Kwok, Lam For, editor

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results