Research map / Vision-language and multimodal / Cross-modal learning
Urban Boundaries, Social Barriers: A Benchmark and Vision-Centric Framework for Mapping Gated Communities and Equity Implications
Minwei Zhao, WEIMING ZHANG, Jiawang DU, Qiming LIU, Weiming Zhuang, Pei Nie, Cai Wu
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 LoRA fine-tuning to the DINOv3-SAT backbone (ablation) | DINOv3-SAT · GBA-GCs Guangzhou | F1: 0.718 -> 0.768 | Improved |
| add BERT-Chinese text metadata (ablation) | DINOv3-SAT + LoRA · GBA-GCs Guangzhou | AUC: 0.850 -> 0.902 | Improved |
| add numerical attributes (ablation) | DINOv3-SAT + LoRA + text · GBA-GCs Guangzhou | F1: 0.813 -> 0.818 | No clear effect |
| add the three-modal CCF fusion block (ablation) | MCGC · GBA-GCs Guangzhou | F1: 0.818 -> 0.833 | Improved |
| add the dual interior-exterior visual stream (ablation) | MCGC · GBA-GCs Guangzhou | F1: 0.833 -> 0.848 | Improved |
| use BERT-Chinese instead of CLIP text encoder (ablation) | MCGC · GBA-GCs Guangzhou | F1: 0.814 -> 0.848 | Improved |
| remove or shuffle text only at test time (ablation) | MCGC · GBA-GCs Guangzhou | F1: 0.848 -> 0.797 | No clear effect |
| remove text in both training and testing (ablation) | MCGC · GBA-GCs Guangzhou | F1: 0.787 -> 0.848 | Improved |
| add IO stream under transfer to Hong Kong (ablation) | MCGC · Hong Kong transfer | accuracy: 0.745 -> 0.863 | Improved |
| add IO stream under transfer to Singapore (ablation) | MCGC · Singapore transfer | AUC: 0.702 -> 0.732 | Mixed |
| multimodal CCF model versus naive fusion on London | MCGC · London transfer | AUC: 0.328 -> 0.627 | Improved |
| corrupt AOI location at test time (ablation) | MCGC · GBA-GCs Guangzhou | F1: .849 -> .797 | Hurt |
Closest papers by meaning
- UV-SAM: Adapting Segment Anything Model for Urban Village Identification (AAAI 2024 · similarity 0.85)
- UrbanVLP: Multi-Granularity Vision-Language Pretraining for Urban Socioeconomic Indicator Prediction (AAAI 2025 · similarity 0.85)
- MotifGPL: Motif-Enhanced Graph Prototype Learning for Deciphering Urban Social Segregation (AAAI 2025 · similarity 0.84)
- CityLens: Evaluating Large Vision-Language Models for Urban Socioeconomic Sensing (ICLR 2026 · similarity 0.84)
- LIVS: A Pluralistic Alignment Dataset for Inclusive Public Spaces (ICML 2025 · similarity 0.84)
- UrbanFeel:A Comprehensive Benchmark for Temporal and Perceptual Understanding of City Scenes through Human Perspective (ICLR 2026 · similarity 0.84)
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.
