The art of picking up signs that a child may be suffering from abuse at home is one of those skills that cannot easily be taught, given its dependence on a range of non-cognitive abilities. It is also difficult to study, given the number of factors that may interfere with this skill in a real-life, professional setting. An immersive virtual reality environment provides a way round these difficulties. In this study, we recruited 64 general practitioners (GPs), with different levels of experience. Would this level of experience have any impact on general practitioners' ability to pick up child-safeguarding concerns? Would more experienced GPs find it easier to pick up subtle (rather than obvious) signs of child-safeguarding concerns? Our main measurement was the quality of the note left by the GP at the end of the virtual consultation: we had a panel of 10 (all experienced in safeguarding) rate the note according to the extent to which they were able to identify and take the necessary steps required in relation to the child safeguarding concerns. While the level of professional experience was not shown to make any difference to a GP's ability to pick up those concerns, the parent's level of aggressive behavior toward the child did. We also manipulated the level of cognitive load (reflected in a complex presentation of the patient's medical condition): while cognitive load did have some impact upon GPs in the "obvious cue" condition (parent behaving particularly aggressively), this effect fell short of significance. Furthermore, our results also suggest that GPs who are less stressed, less neurotic, more agreeable and extroverted tend to be better at raising potential child abuse issues in their notes. These results not only point at the considerable potential of virtual reality as a training tool, they also highlight fruitful avenues for further research, as well as potential strategies to support GP's in their dealing with highly sensitive, emotionally charged situations.

Figure 1: Overview of the two prototypes of collaborative applications presented in this paper that are based on Ubiq-Genie. Left: a voice-controlled texture generation method based on the diffusion-based image synthesis model Stable Diffusion 2.0 [21]. Right: a voice-controlled conversational agent based on the text synthesis tool ChatGPT [19].

Figure 1: Conceptual overview of our outdoor CMR system showing a detailed 3D model of a street that a local AR user can experience in collaboration with a remote VR user. The virtual coordinate space of the shared virtual environment is aligned to the real-world surroundings of the AR user (Space B) with QR code markers. Line illustrations by Suhyun Park (artist).