The field of urban planning and image generation is witnessing significant advancements, driven by the integration of machine learning models and high-resolution data. Researchers are developing innovative approaches to assess walkability, a crucial factor in promoting physical activity and public health. These efforts involve creating standardized, high-resolution walkability indices that can be applied across diverse urban contexts. In the realm of image generation, studies are focusing on optimizing multi-round enhanced training in diffusion models to improve preference understanding and image consistency. Furthermore, there is a growing concern about privacy leakage through image geolocation, highlighting the need for privacy-aware development in agentic multi-modal large reasoning models. Noteworthy papers include:

• A study on developing a high-resolution walkability index for Europe, which provides a practical tool for researchers and policymakers.
• A paper on optimizing multi-round enhanced training in diffusion models, which significantly improves image consistency and alignment with user intent.
• A research work on revealing privacy leakage in image geolocation for agentic multi-modal large reasoning models, which emphasizes the urgent need for privacy-aware development.