The field of urban and ecological informatics is witnessing significant developments, driven by the increasing availability of large datasets and advancements in artificial intelligence and deep learning techniques. Researchers are leveraging these technologies to automate tasks such as species identification, building heritage assessment, and urban tree biodiversity mapping, which were previously labor-intensive and required expert knowledge.

A notable trend is the integration of multi-modal data sources, including street-level imagery, satellite imagery, and sensor data, to produce structured urban spatial information and support data-driven decision-making. The use of weakly supervised and unsupervised learning approaches is also gaining traction, enabling the estimation of biodiversity and tree instance segmentation without the need for extensive labeled datasets.

Particularly noteworthy are the studies that have demonstrated the potential of lightweight deep learning models for real-time species identification and the development of frameworks for assessing streetscape inclusivity and site planning layout indicators. For example, one study achieved an average accuracy of 99.29% in classifying wood species using a custom image dataset and a convolutional neural network architecture. Another study presented a mixed-methods approach that combined participatory research with AI-based analysis to assess streetscape inclusivity, revealing variations in perceptions of inclusivity and accessibility across demographic groups.