Bo Qiu

I am currently a third-year M.S. student at LIESMARS, Wuhan University, under the supervision of Prof. Zhen Dong and Prof. Bisheng Yang. Before this, I received the B.S. degree from the School of Geodesy and Geomatics, Wuhan University, in 2022.

My research interests include 3D scene understanding, point cloud processing, and their applications in digital twin cities and intelligent transportation systems.


Education
  • LIESMARS, Wuhan University
    LIESMARS, Wuhan University
    M.S. in Photogrammetry and Remote Sensing
    Sep. 2022 - Now
  • SGG, Wuhan University
    SGG, Wuhan University
    B.S. in Geodesy and Geomatics Engineering
    Sep. 2018 - Jun. 2022
Honors & Awards
  • First Prize Scholarship (Top 5%), Wuhan University, Three Times
    2024, 2023, 2019
  • Outstanding Graduate Student (Top 10%), Wuhan University
    2024, 2023
  • The Best Paper Award (Top 10) at the 7th National LiDAR Conference
    2023
  • Grand Prize (Ranking 1st) in Data Processing Competition of the 7th National LiDAR Conference
    2023
  • 2nd Place Award in 3D Semantic Segmentation to the Open World (3DOW) Challenge, IROS
    2022
  • Outstanding Graduates (Top 5%), Wuhan University
    2022
  • Merit Student (Top 10%), Wuhan University
    2019
  • Hi-Target Special Scholarship (Top 10%), Wuhan University
    2019
  • Outstanding Student (Top 20%), Wuhan University, Twice
    2021, 2020
  • Second Prize Scholarship (Top 20%), Wuhan University
    2020
  • Second Prize in MathorCup College Mathematical Modeling Challenge
    2021
  • National Second Prize in the Chinese Mathematics Competitions
    2019
  • Excellent Student Cadre, Wuhan University
    2019
  • Advanced Individual of Social Work, Wuhan University
    2019
  • Outstanding Volunteer, Wuhan University
    2019
  • First Prize for Outstanding Achievements in Volunteer Service, Wuhan University
    2019
Selected Publications (view all )
WHU-Railway3D: A Diverse Dataset and Benchmark for Railway Point Cloud Semantic Segmentation
WHU-Railway3D: A Diverse Dataset and Benchmark for Railway Point Cloud Semantic Segmentation

Bo Qiu*, Yuzhou Zhou*, Lei Dai, Bing Wang, Jianping Li, Zhen Dong†, Chenglu Wen, Zhiliang Ma, Bisheng Yang (* equal contribution)

IEEE TITS (In Press, SCI, JCR Q1, IF=7.9, TOP Journal in Engineering Technology) 2024

This paper introduces WHU-Railway3D, a diverse point cloud semantic segmentation (PCSS) dataset specifically tailored for railway scenes. Spanning approximately 30 km and comprising 4.6 billion points, the dataset includes 11 richly annotated categories across urban, rural, and plateau railway environments.

WHU-Railway3D: A Diverse Dataset and Benchmark for Railway Point Cloud Semantic Segmentation
WHU-Railway3D: A Diverse Dataset and Benchmark for Railway Point Cloud Semantic Segmentation

Bo Qiu*, Yuzhou Zhou*, Lei Dai, Bing Wang, Jianping Li, Zhen Dong†, Chenglu Wen, Zhiliang Ma, Bisheng Yang (* equal contribution)

IEEE TITS (In Press, SCI, JCR Q1, IF=7.9, TOP Journal in Engineering Technology) 2024

This paper introduces WHU-Railway3D, a diverse point cloud semantic segmentation (PCSS) dataset specifically tailored for railway scenes. Spanning approximately 30 km and comprising 4.6 billion points, the dataset includes 11 richly annotated categories across urban, rural, and plateau railway environments.

Ubiquitous Point Cloud: Theory, Model, and Applications (1st ed.)
Ubiquitous Point Cloud: Theory, Model, and Applications (1st ed.)

Bisheng Yang, Zhen Dong, Fuxun Liang, Xiaoxin Mi, et al.

CRC Press 2024

This book provides the latest theory and methodology for point cloud processing with Al to better serve earth observation, 3D vision, autonomous driving, smart city, and geospatial information applications. (I was responsible for writing Chapter 7 of this book, titled Point Cloud Semantic Segmentation.)

Ubiquitous Point Cloud: Theory, Model, and Applications (1st ed.)
Ubiquitous Point Cloud: Theory, Model, and Applications (1st ed.)

Bisheng Yang, Zhen Dong, Fuxun Liang, Xiaoxin Mi, et al.

CRC Press 2024

This book provides the latest theory and methodology for point cloud processing with Al to better serve earth observation, 3D vision, autonomous driving, smart city, and geospatial information applications. (I was responsible for writing Chapter 7 of this book, titled Point Cloud Semantic Segmentation.)

Superpoint-Assisted Neural Radiance Fields for Point Cloud Semantic Segmentation without 3D Annotation
Superpoint-Assisted Neural Radiance Fields for Point Cloud Semantic Segmentation without 3D Annotation

Zhen Cao*, Xiaoxin Mi*, Bo Qiu, Zhipeng Cao, Chen Long, Xinrui Yan, Chao Zheng, Zhen Dong†, Bisheng Yang (* equal contribution)

ISPRS Journal of Photogrammetry and Remote Sensing (Under Review, SCI, JCR Q1, IF=12.7, TOP Journal in Geosciences) 2024

This study introduces a superpoint-assisted Neural Radiance Field (NeRF) for point cloud semantic segmentation without requiring additional training data. For a given scene, the proposed method uses point clouds, associated images, and image pseudo-semantic labels as inputs to achieve pointwise semantic segmentation.

Superpoint-Assisted Neural Radiance Fields for Point Cloud Semantic Segmentation without 3D Annotation
Superpoint-Assisted Neural Radiance Fields for Point Cloud Semantic Segmentation without 3D Annotation

Zhen Cao*, Xiaoxin Mi*, Bo Qiu, Zhipeng Cao, Chen Long, Xinrui Yan, Chao Zheng, Zhen Dong†, Bisheng Yang (* equal contribution)

ISPRS Journal of Photogrammetry and Remote Sensing (Under Review, SCI, JCR Q1, IF=12.7, TOP Journal in Geosciences) 2024

This study introduces a superpoint-assisted Neural Radiance Field (NeRF) for point cloud semantic segmentation without requiring additional training data. For a given scene, the proposed method uses point clouds, associated images, and image pseudo-semantic labels as inputs to achieve pointwise semantic segmentation.

RailSeg: Learning Local–Global Feature Aggregation With Contextual Information for Railway Point Cloud Semantic Segmentation
RailSeg: Learning Local–Global Feature Aggregation With Contextual Information for Railway Point Cloud Semantic Segmentation

Tengping Jiang, Bisheng Yang†, Yongjun Wang†, Lei Dai, Bo Qiu, Shan Liu, Shiwei Li, Qinyu Zhang, Xin Jin, Wenjun Zeng

IEEE TGRS (SCI, JCR Q1, IF=7.5) 2023

This study introduces RailSeg, a deep learning framework for railway point cloud semantic segmentation, focusing on point cloud downsampling, integrated local–global feature extraction, spatial context aggregation, and semantic regularization.

RailSeg: Learning Local–Global Feature Aggregation With Contextual Information for Railway Point Cloud Semantic Segmentation
RailSeg: Learning Local–Global Feature Aggregation With Contextual Information for Railway Point Cloud Semantic Segmentation

Tengping Jiang, Bisheng Yang†, Yongjun Wang†, Lei Dai, Bo Qiu, Shan Liu, Shiwei Li, Qinyu Zhang, Xin Jin, Wenjun Zeng

IEEE TGRS (SCI, JCR Q1, IF=7.5) 2023

This study introduces RailSeg, a deep learning framework for railway point cloud semantic segmentation, focusing on point cloud downsampling, integrated local–global feature extraction, spatial context aggregation, and semantic regularization.

All publications
Selected Projects (view all )
Detailed Extraction of Roadside Trees from MLS Point Cloud
Detailed Extraction of Roadside Trees from MLS Point Cloud

Guided by Professor Bisheng Yang and Zhen Dong from Wuhan University

2023.06 - 2023.10

As one of the core developers. We proposed a confidence-guided roadside tree extraction algorithm. The core idea is to first segment pole-like objects and calculate their confidence of being trees, then classify these pole-like targets into trees and non-trees. Next, we use a combination of weighted distance constraints and an allometric model to extract individual trees using the optimized minimum cut method. Based on the extraction results of individual trees, we estimate tree parameters and ultimately obtain 3D tree models through 3D modeling techniques.

Detailed Extraction of Roadside Trees from MLS Point Cloud
Detailed Extraction of Roadside Trees from MLS Point Cloud

Guided by Professor Bisheng Yang and Zhen Dong from Wuhan University

2023.06 - 2023.10

As one of the core developers. We proposed a confidence-guided roadside tree extraction algorithm. The core idea is to first segment pole-like objects and calculate their confidence of being trees, then classify these pole-like targets into trees and non-trees. Next, we use a combination of weighted distance constraints and an allometric model to extract individual trees using the optimized minimum cut method. Based on the extraction results of individual trees, we estimate tree parameters and ultimately obtain 3D tree models through 3D modeling techniques.

Point2Models Software Development
Point2Models Software Development

Guided by Professor Bisheng Yang and Zhen Dong from Wuhan University

2022.09 - 2023.09

As one of the developers and algorithm maintainers, I developed and maintained urban MLS point cloud object extraction algorithms, point cloud coordinate transformation and projection algorithms, and DEM processing algorithms.

Point2Models Software Development
Point2Models Software Development

Guided by Professor Bisheng Yang and Zhen Dong from Wuhan University

2022.09 - 2023.09

As one of the developers and algorithm maintainers, I developed and maintained urban MLS point cloud object extraction algorithms, point cloud coordinate transformation and projection algorithms, and DEM processing algorithms.

GIS Software Tutorial Writing Team
GIS Software Tutorial Writing Team

Guided by Associate Professor Yuejin Deng from Wuhan University

2020.09 - 2021.08

As a team member, I studied and organized 3D visualization materials related to Cesium and was responsible for writing the book titled '3D Visualization Development Guide Based on Cesium.'

GIS Software Tutorial Writing Team
GIS Software Tutorial Writing Team

Guided by Associate Professor Yuejin Deng from Wuhan University

2020.09 - 2021.08

As a team member, I studied and organized 3D visualization materials related to Cesium and was responsible for writing the book titled '3D Visualization Development Guide Based on Cesium.'

Visual Localization and Navigation Methods for Quadruped Robot Platforms
Visual Localization and Navigation Methods for Quadruped Robot Platforms

Guided by Professor Chi Guo from Wuhan University

2020.04 - 2021.10 National Undergraduate Training Program for Innovation and Entrepreneurship (No. 202010486087)

As a core researcher and developer, I was primarily responsible for the implementation and integration of visual SLAM (Simultaneous Localization and Mapping) algorithms, while also contributing to the research on a binocular vision-based system for object detection, tracking, and trajectory prediction.

Visual Localization and Navigation Methods for Quadruped Robot Platforms
Visual Localization and Navigation Methods for Quadruped Robot Platforms

Guided by Professor Chi Guo from Wuhan University

2020.04 - 2021.10 National Undergraduate Training Program for Innovation and Entrepreneurship (No. 202010486087)

As a core researcher and developer, I was primarily responsible for the implementation and integration of visual SLAM (Simultaneous Localization and Mapping) algorithms, while also contributing to the research on a binocular vision-based system for object detection, tracking, and trajectory prediction.

All projects