profile photo

Huaqi Tao

Huaqi is currently a Master's student in the Robotics and Computer Vision Lab, Southern University of Science and Technology (SUSTech), advised by Chair Professor Hong Zhang.

Prior to that, he received his B.Eng. in Mechanical Engineering from Guangxi University. He has also interned at Institute of Automation, Chinese Academy of Sciences (CASIA).

Looking for a PhD position starting in Fall 2027.

Email  /  CV  /  Google Scholar  /  GitHub

Research Vision

I aim to build intelligent robotic navigation systems that enable robots to answer three fundamental questions: Where am I? Where should I go? How can I get there? My research focuses on visual navigation, where robots rely on visual perception to understand their surroundings, localize themselves, reason about goals, and move safely through complex environments.

I am interested in navigation systems that go beyond predefined maps and rigid pipelines. Instead, I seek to develop robots that can use visual perception, spatial understanding, and language-guided reasoning to navigate in a more human-like and adaptive manner.

To that end, I focus on:

  • Visual place recognition and visual relocalization, enabling robots to identify where they are and recover their position from visual observations;
  • Scene-graph-based and VLM-driven vision-language navigation, allowing robots to understand semantic spatial relationships and follow language instructions in complex environments;
  • Efficient local representations for planning and obstacle avoidance, supporting real-time motion planning and safe navigation in dynamic or cluttered scenes.

I believe the future of robot navigation lies in enabling robots to perceive and act in the world as embodied intelligent agents. My work is an attempt to build navigation systems that do not simply compute paths, but that understand space, reason about goals, and move through the world with increasing autonomy, adaptability, and intelligence.

Publications

Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting
Huaqi Tao, Bingxi Liu, Guangcheng Chen, Fulin Tang, Li He, Hong Zhang
CVPR, 2026

paper  /  project page  /  code

TextInPlace: Indoor Visual Place Recognition in Repetitive Structures with Scene Text Spotting and Verification
Huaqi Tao, Bingxi Liu, Calvin Chen, Tingjun Huang, He Li, Jinqiang Cui, Hong Zhang
IROS, 2025

paper  /  code

A Convolutional-Transformer Network for Crack Segmentation with Boundary Awareness
Huaqi Tao, Bingxi Liu, Jinqiang Cui, Hong Zhang
ICIP, 2023

paper  /  code