Education
Ph.D. in Computer Science and Technology
Beihang University · Beijing
Beijing ↔ Shanghai2026
About
Peizhou Cao.
Building models and worlds that help machines understand space.
I am a second-year Ph.D. student at Beihang University and Shanghai AI Laboratory, co-supervised by Jiangmiao Pang and Dahua Lin. My work focuses on spatial intelligence, multimodal foundation models, and world models for understanding and simulating visual environments.
01 / Profile
Experience
Research
Research Intern
Shanghai AI Laboratory · Shanghai
Education
B.Eng. in Electronic Engineering
Xidian University · Xi’an
2021 — 2022 National Scholarship
Spatial perceptionVision-language reasoningWorld modelsScalable simulation
02 / Research
Research landscape
Models, data, benchmarks, and integrated systems across my work.
Layer 01 · CoreSpatial intelligencePerceive → reason → simulate
Layer 03 · Model papers
G²TAMGeometry-grounded tracking
Thinking with ImaginationAgentic spatial reasoning
PSDMSemi-distillation model
Layer 03 · Benchmark papers
PerceptionBenchAtomic visual perception
MMSI-Video-BenchVideo spatial intelligence
Layer 03 · Integrated projects
Kimi K3Multimodal foundation model
GRUtopiaEmbodied simulation platform
Research architecture
Spatial intelligence
Papers are organized by their primary role in the research pipeline.
Model
Data
Benchmark
Integrated systems
03 / Selected work
Research, built to be used.
Nine selected works spanning models, benchmarks, datasets, and large-scale intelligent systems.
G²TAM: Geometry Grounded Track Anything Model
Paper ↗
Code coming soon
Progressive semi-distillation
PAN + MS · progressive guidance · cooperation + correction