The authors formulate a new setting: grounded urban world construction from sparse aerial observations, where the goal is to jointly recover instance-level building geometry, explicit road topology, and detailed appearance without repeated dense reconstruction.
Sparse aerial photos become persistent, agent-editable 3D city models
Sun et al. show how one low-altitude aerial RGB image per region can yield metrically scaled buildings, explicit road connectivity, and versioned city states that agents can query and edit over long task chains.
Chinese Tech
Jiahui Sun · Ran Bi · Yifei Liu · Xiaolong Zhou · Yishuo Xu · Yuanqi Li · +5 more
Nanjing University · Shanghai Artificial Intelligence Laboratory · University of Science and Technology of China · SenseTime · Shanghai Jiao Tong University
Research Digest··3 min read
Sun et al.
Why this paper
From SenseTime and 5 others
In one line
Sparse aerial observations can be converted into persistent, agent-operable 3D cities with metric geometry, road connectivity, and editable entity identities.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§