Tong Nie

About

Tong Nie is a rising researcher in computer vision and robotics, with a primary focus on 3D perception and object pose estimation. His most influential work, "Category-Level 6D Pose Estimation Using Geometry-Guided Instance-Aware Prior and Multi-Stage Reconstruction" (2023), addresses a critical challenge in robotic manipulation and augmented reality: accurately predicting the translation and rotation of unseen object instances from known categories without requiring per-instance 3D models. By introducing a geometry-guided instance-aware prior combined with a multi-stage reconstruction pipeline, Nie’s method significantly improves generalization across diverse object shapes, enabling robust 6D pose estimation in real-world scenarios. This contribution has already garnered 8 citations, reflecting its timely relevance to the field. Nie’s research bridges the gap between category-level understanding and precise geometric reasoning, offering practical solutions for autonomous systems that must interact with novel objects. His work is particularly notable for its potential to advance embodied AI, where reliable pose estimation is foundational for tasks like grasping and scene manipulation. As a young scholar, Tong Nie is establishing himself as a key contributor to the next generation of 3D vision technologies.

Research Focus

Key Achievements

Top Papers

1. 1
   Category-Level 6D Pose Estimation Using Geometry-Guided Instance-Aware Prior and Multi-Stage Reconstruction

   8 citations · 2023

Key Collaborators

Contact & Links