CommuniWave A Machine Learning Model for Quantifying the Degree of Temporary Informal Behavior in Urban Communities

Jul 16, 2026

Upload by Sage

Urban managers and designers need practical ways to strengthen territorial resilience amid complexity and uncertainty. Community planning is often top-down and lacks effective metrics to quantify residents’ temporary informal behaviors, which can lead to repeated conflicts with original plans. CommuniWave proposes a machine learning pipeline to detect and quantify the Degree of Informal Behavior (DIB). It combines a Behavior Capture Net (BCN) from mmaction2, a YOLOv10-based detector, and a random-forest evaluation model, generating DIB fluctuation charts from street videos to enable dynamic monitoring and refined decision-making.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

Introduction

Territorial resilience and informal resident activities

Origins and manifestations of informal behavior

Prior research on planning, governance, and evaluation

Computational and machine-learning approaches

Video-based high-precision behavior analysis

COMMUNIWAVE A Machine Learning Model for Quantifying the Degree of Temporary Informal Behavior in Urban Communities HONGYE YANG, SHIEN LIU Beijing Institute of Architectural Design Co., Ltd [yanghongye@biad.com.cn](yanghongye@biad.com.cn) , [hyang783@gatech.edu](hyang783@gatech.edu) Wuxi Taihu University [212017075@wxu.edu.cn](212017075@wxu.edu.cn) AND ZHIHAO XIE Wuxi Taihu University [212017074@wxu.edu.cn](212017074@wxu.edu.cn) Abstract. For urban managers and designers, improving the functional attributes of urban communities to enhance territorial resilience in the face of complexity and uncertainty is crucial. Currently, community planning often follows a top-down approach and lacks effective metrics to quantify informal behaviors of residents, leading to frequent conflicts with original plans. This study introduces CommuniWave, a machine learning model designed to efficiently detect and quantify the Degree of Informal Behavior (DIB) in urban communities. The model integrates a Behavior Capture Net (BCN) based on mmaction2, a self-developed YOLOv10 model (YLX), and a Behavior Eval Model (BEM) using random forest. Ultimately, by generating DIB fluctuation charts from street videos, the model facilitates dynamic monitoring, supporting urban managers in making refined decisions to enhance the overall resilience of communities. Keywords: Machine Learning, Spatio Temporal Action Detection, Urban Communities, Degree of Informal Behavior, Refined Design 1. Introduction Territorial resilience is an interdisciplinary concept of sustainable development(Brunetta et al., 2019) , emphasizing the need for decisionmakers to adopt flexible and proactive approaches in addressing risks and challenges within a given space. In urban communities, spontaneous bottomup activities by residents represent uncertain factors. However, management approaches toward these activities often lack sustainable understanding. Previous studies have shown that a community management model combining bottom-up and top-down approaches helps reduce conflicts, thereby promoting sustainable territorial resilience(Semeraro et al., 2020) . Informal behavior in urban communities is an informal phenomenon whose research origins can be traced back to economist Keith Hart's discussion of the Informal Sector(Hart, 1985) . In communities, the disconnect between government-led urban planning and the daily needs of residents is the main reason for the emergence of informal behaviors(Anon., 2018) . These behaviors often manifest as spontaneous, context-specific responses, such as randomly appearing street stalls or spontaneously formed gathering places(Chase, Crawford and Kaliski, 2008) . These informal behaviors of community residents often accompany rapid urban expansion and unbalanced land policies, reflecting issues of imbalance between infrastructure supply and demand(Frederic Deng and Huang, 2004) . Scholars generally believe that these spontaneous activities have positive significance for community life. Jane Jacobs, in The Death and Life of Great American Cities, emphasized that urban planning should create conditions for “casual public interactions” in community life(Jacobs, 1992); Jan Geh pointed out that the vitality of street spaces depends on social interactions in daily life(Gehl, 1971); Margaret Crawford argued that urban managers need to understand and accept the visual "disorder" brought about by the"counter-publics" actions of street vendors and the homeless(Crawford, 1995) . Through residents' life experiences, these behaviors continuously reshape and redefine public spaces and domains, promoting the diversity and vitality of neighborhood activities(Mehta and Bosson, 2021) . In contemporary society, urban managers have begun to try to guide these bottom-up spontaneous behaviors within a compliant framework (Yao et al., 2024) . In Beijing, China, dedicated community duty planners (CDPs) have been introduced to coordinate neigh

FAQs

What problem does CommuniWave address in urban community planning?

CommuniWave targets the lack of effective quantitative metrics for residents’ temporary informal behaviors in top-down community planning, which often results in conflicts with original plans.

How does CommuniWave detect and quantify the Degree of Informal Behavior (DIB)?

The model integrates a Behavior Capture Net (BCN) built on mmaction2, a self-developed YOLOv10-based detector (YLX), and a Behavior Eval Model (BEM) using random forest to compute DIB.

How can urban managers use CommuniWave outputs in practice?

By producing DIB fluctuation charts from street videos, the approach supports dynamic monitoring and refined decision-making to improve community resilience.

Jul 16, 2026

Upload by Sage