CompetitorFormer: Mitigating Query Conflicts for 3D Instance Segmentation via Competitive Strategy
动态 News
- 2026.08 PointQ-Bench 被 ACM MM 2026 接收。 PointQ-Bench is accepted by ACM MM 2026.
- 2026.08 Boreal3D 发表于 Remote Sensing of Environment。 Boreal3D is published in Remote Sensing of Environment.
- 2026.02 CompetitorFormer 被 CVPR 2026 接收。 CompetitorFormer is accepted by CVPR 2026.
- 2025.10 一篇三维实例分割论文发表于 IEEE Robotics and Automation Letters。 One paper on 3D instance segmentation is published in IEEE Robotics and Automation Letters.
- 2025 一篇树点云叶木分离论文发表于 IEEE Transactions on Geoscience and Remote Sensing。 One paper on leaf–wood separation in tree point clouds is published in IEEE Transactions on Geoscience and Remote Sensing.
教育经历 Education
- 博士研究生,西安交通大学软件学院,导师王迪副教授 Ph.D. student, School of Software Engineering, Xi'an Jiaotong University, supervised by Assoc. Prof. Di Wang
- 硕士,西安电子科技大学,导师王迪副教授 M.S., Xidian University, supervised by Assoc. Prof. Di Wang
- 学士,武汉科技大学 B.S., Wuhan University of Science and Technology
论文 Publications
已接收 / 发表 9 篇 · 预印本 2 篇 Accepted / published: 9 · Preprint: 2
PointQ-Bench: Benchmarking Diagnostic and Interpretable Point Cloud Quality Assessment
A synthetic data generation framework for deep learning-based LiDAR forest structure analysis
VertiCue-Bench: Diagnosing Whether MLLMs Use Height Cues to Resolve 2D Ambiguity in Remote Sensing Natural Scenes
ForestHG-Trace: Traceable Long-Horizon Ecological Reasoning over Large-Scale Forest Scenes
Enhancing 3D Instance Segmentation with Dense Connection Decoder and Layer-Aware Fusion
Optimizing Label Efficiency for Learning-Based Leaf–Wood Separation in Tree Point Clouds
Towards the Digital Twin of Urban Forest: 3D Modeling and Parameterization of Large-Scale Urban Trees from Close-Range Laser Scanning
Gas Source Localization using Dueling Deep Q-Network with an Olfactory Quadruped Robot
Multi-sensory Olfactory Quadruped Robot for Odor Source Localization
An Active Pursuit Strategy for Autonomous Mobile Robots based on Deep Reinforcement Learning
* 表示共同一作。姓名加粗为本人。 * Equal contribution. My name is in bold.