Video conferencing allows distance-separated family members to interact somewhat akin to being together at the same place and time. Yet most video conferencing systems are designed for phone-like calls between only two locations. Using such systems for long interactions or social gatherings with multiple families is cumbersome, if not impossible. For this reason, we wanted to explore how families would make use of a video system that permitted sharing everyday life over extended periods of time between multiple locations. We designed a media space called Family Portals that provides shared video between three locations and deployed it within the homes of six families. Results show that the media space increased feelings of connectedness and the focus on a triad, in contrast to a dyad, caused new styles of interaction to emerge. Despite this, families experienced new privacy challenges and non-adoption by some family members, not previously seen in dyadic family media spaces.

One important problem in computer vision is to provide a demographic description a person from an image. In practice, many of the state-of-the-art methods use only an analysis of the face to estimate the age and gender of a person of interest. We present a model that combines two problems, height estimation and demographic classification, which allows each to serve as context for the other. Our idea is to use a calibrated camera for measuring the height of people in the scene. Height is measured by jointly inferring across anthropometric dimensions, age, and gender using publicly available statistics. The height estimate provides context for recognizing the age and gender of the subject, and likewise age and gender conditions the distribution of the anthropometric features for estimating height. The performance of our method is explored on a new database of 127 people captured with a calibrated camera with recorded height, age, and gender. We show that estimating height leads to improvements in age and gender classification, and vice versa. To the best of our knowledge, our model produces the most accurate automatic height estimates reported, with the error having a standard deviation of 26.7 mm.