Vision-language models offer scalable solution for urban blight assessment

Researchers have developed a new framework to assess urban blight using large vision-language models, offering a scalable and cost-effective alternative to traditional manual surveys. By analyzing multiple street views and housing attributes like roof integrity and wall damage, these models can provide binary assessments and probabilistic estimates of disrepair. An ensemble approach combining XGBoost with weighted scoring demonstrated superior performance and robustness compared to individual models, enabling low-cost tracking and management of housing stock conditions. AI

IMPACT This research demonstrates a novel application of vision-language models for urban planning and management, potentially improving efficiency and cost-effectiveness in blight assessment.

RANK_REASON Academic paper detailing a new methodology and its evaluation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• Detroit
• Hugging Face
• XGBoost

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →