About

Yiyi Liao is a prominent researcher at the intersection of computer vision, robotics, and 3D scene understanding, whose work has significantly advanced autonomous perception systems. Best known for co-creating **KITTI-360**, a landmark dataset for urban scene understanding in 2D and 3D, Liao helped bridge the historically siloed fields of computer vision, graphics, and robotics — a contribution that has garnered over 630 citations and become an essential benchmark for self-driving car research worldwide. Liao's early work demonstrated a keen interest in making robots smarter with limited sensor data. Her 2017 paper on monocular depth estimation using partial laser observation (126 citations) showed how standard robotic platforms could infer 3D geometry without dedicated depth sensors — a practically impactful contribution for resource-constrained systems. Her scene classification research using convolutional neural networks (103 citations) further established her expertise in semantic scene understanding. More recently, Liao has pushed into neural implicit representations for large-scale LiDAR mapping (NF-Atlas) and robust global localization with RING++, reflecting her evolving focus on scalable, real-world robotic mapping. Across diverse topics — from visual odometry to traversable region detection — her body of work consistently targets the grand challenge of enabling machines to understand and navigate complex physical environments.

Research Focus

Key Achievements

8
H-Index
12
Papers
1,042
Total Citations
87
Avg Citations/Paper
🏆 Most Cited Paper
KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3D
630 citations · 2022
📈 Most Prolific Year: 2016 (4 Papers)
🤝 Key Collaborators: 30
🏛 Institutions: TH Bingen University of Applied Sciences, Zhejiang University, State Key Laboratory of Industrial Control Technology, Zhejiang University of Technology

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago