“UrbanBIS: A Large-scale Benchmark for Fine-grained Urban Building Instance Segmentation” by Yang, Xue, Zhang, Xie, Fu, et al. …
Conference:
Type(s):
Title:
- UrbanBIS: A Large-scale Benchmark for Fine-grained Urban Building Instance Segmentation
Session/Category Title: Procedural Modeling
Presenter(s)/Author(s):
Moderator(s):
Abstract:
UrbanBIS is a large-scale urban understanding benchmark, with six scenes, 2.5 billion points, and 3,370 buildings, supporting urban-level semantic and instance segmentation. It also introduces building sub-categories and multi-mode data, facilitating various intelligent city applications. We propose B-Seg, achieving better accuracy with faster inference speed on instance segmentation on UrbanBIS.
Overview Page:
Submit a story:
If you would like to submit a story about this presentation, please contact us: historyarchives@siggraph.org
