When producing an engaging augmented reality (AR) user experience, it is crucial to create AR content that mimics real-life objects’ behavior to the greatest extent possible. A critical aspect to achieve this is ensuring that the digital objects conform to line-of-sight rules and are either partially or completely occluded, just like real-world objects would be. The study explores the concept of utilizing a pre-existing 3D representation of the physical environment as an occlusion mask that governs the rendering of each pixel. Specifically, the research aligns a Level of Detail (LOD) 1 building model and a 3D mesh model with their real-world counterparts and evaluates the effectiveness of occlusion between the two models in an outdoor setting. Despite the mesh model containing more detailed information, the overall results do not show improvement. In an indoor scenario, the researchers leverage the scanning capability of HoloLens 2.0 to create a pre-scanned representation, which helps overcome the limited range and delay of the mesh reconstruction.

Determining emotions of people during different activities (e.g. surfing the web or walking and driving in urban areas) is of high interest to many industries such as the advertising and marketing industry (Imotions, 2022) or city developers and planners (Zeile et al., 2015). However, as the authors have explained in their previous works (Schneider et al., 2020, Dastageeri et al., 2019, Kohn et al., 2018), it is very difficult to determine emotions using surrogate measurements. This is compounded by the fact that many people have trouble to name their emotions correctly as they are often mixed and hardly ever occur as a single and distinct emotion. In order to improve the attractiveness of cities not just based on general presumptions about how citizens would react to certain changes in the urban environment, but based on physical measurements, previous works have shown approaches to do that (Schneider et al., 2020, Dastageeri et al., 2019). They have developed a first attempt to correlate measured physical parameters such as heart rate and skin conductivity (among others) which are triggered by location and environment to emotional states using machine learning. To correlate locations to emotions is an important aspect for city planners, as a person’s emotion for a location defines the personal relationship to a place which can help to gauge the attractiveness of a place and give indicators about where to improve the city or place.