Research map / 3D vision / 3D generation
Urban Zoning Using Higher-Order Markov Random Fields on Multi-View Imagery Data
Tian Feng, Quang-Trung Truong, Duc Thanh Nguyen, Jing Yu Koh, Lap-Fai Yu, Alexander Binder, Sai-Kit Yeung
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add street-view photo higher-order potentials to satellite-only zoning (ablation) | MS-CNN · Urban zoning, New York | zoning accuracy: 66.70% -> 70.33% | Improved |
| add street-view photo higher-order potentials to satellite-only zoning (ablation) | MS-CNN · Urban zoning, San Francisco | zoning accuracy: 64.68% -> 71.80% | Improved |
| add street-view photo higher-order potentials to satellite-only zoning (ablation) | MS-CNN · Urban zoning, Boston | zoning accuracy: 51.33% -> 57.09% | Improved |
| add satellite imagery to street-view-only zoning (ablation) | HO-MRF · Urban zoning, San Francisco | zoning accuracy: 61.63% -> 71.80% | Improved |
| replace conventional MRF with higher-order MRF (ablation) | MRF · Urban zoning, New York | zoning accuracy: 66.87% -> 70.33% | Improved |
| replace conventional MRF with higher-order MRF (ablation) | MRF · Urban zoning, San Francisco | zoning accuracy: 62.99% -> 71.80% | Improved |
| replace conventional MRF with higher-order MRF (ablation) | MRF · Urban zoning, Boston | zoning accuracy: 47.95% -> 57.09% | Improved |
| use the proposed MS-CNN instead of the adapted Volpi and Tuia network (ablation) | satellite pixel classifier · Satellite pixel classification, New York | pixel accuracy: 20.50% -> 66.70% | Improved |
| use Places-CNN features instead of GIST with random forest (ablation) | street-view photo classifier · Street-view photo classification, New York (overall) | accuracy: 67.51% -> 79.76% | Improved |
| use Places-CNN features instead of HOG with random forest (ablation) | street-view photo classifier · Street-view photo classification, Boston (overall) | accuracy: 71.77% -> 82.08% | Improved |
| use random forest instead of SVM-RBF on Places-CNN features (ablation) | street-view photo classifier · Street-view photo classification, New York (overall) | accuracy: 72.33% -> 79.76% | Improved |
| use random forest instead of a fully connected network on Places-CNN outputs (ablation) | street-view photo classifier · Street-view photo classification, New York (overall) | accuracy: 60.89% -> 79.76% | Improved |
Closest papers by meaning
- UrbanDataLayer: A Unified Data Pipeline for Urban Science (NeurIPS 2024 · similarity 0.82)
- Urban Region Embedding via Multi-View Contrastive Prediction (AAAI 2024 · similarity 0.82)
- UV-SAM: Adapting Segment Anything Model for Urban Village Identification (AAAI 2024 · similarity 0.82)
- UrbanVLP: Multi-Granularity Vision-Language Pretraining for Urban Socioeconomic Indicator Prediction (AAAI 2025 · similarity 0.81)
- UrbanFeel:A Comprehensive Benchmark for Temporal and Perceptual Understanding of City Scenes through Human Perspective (ICLR 2026 · similarity 0.81)
- Improving Region Representation Learning from Urban Imagery with Noisy Long-Caption Supervision (AAAI 2026 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
