Public adoption of camera-equipped mobile phones has given the average observer of an event the ability to capture their perspective and upload the video for online viewing (e.g. YouTube). When traditional wide-area surveillance systems fail to capture an area or time of interest, crowd-sourced videos can provide the information needed for event reconstruction. This paper presents the first end-to-end method for automatic cross-camera tracking from crowd-sourced mobile video data.Our processing (1) sorts videos into overlapping space-time groups, (2) finds the inter-camera relationships from objects within each view, and (3) provides an end user with multiple stabilized views of tracked objects. We demonstrate the system's effectiveness on a real dataset collected from YouTube.

No abstract

What are the social bases of neighborhood formation in urban areas, and at what spatial scale are they most distinct from other neighborhoods? We address these questions in the case of St. Louis, Missouri, in 1930, where we can take advantage of unique geocoded census microdata on the whole population of the city that identifies who, with what background characteristics, lived where. Our analyses show that homophily by race and ethnicity was by far the strongest factor linking characteristics of persons to the composition of their neighbors. Measures of social class also were quite important, while the person’s nativity and family status were statistically significant but minor predictors. Yet while this hierarchy of social factors held for the population as a whole, their relative importance varied greatly across racial/ethnic groups. Similarity in social class to neighbors was most important for native whites, nativity counted as much or more than class for recently arriving immigrant groups including Russians, Italians, and Poles, and race/ethnicity was by far the key predictor for these groups and blacks. We also found that these patterns of homophily were clearest at the scale of individual street segment and first-order combinations of segments. They were similar but less distinct at a larger spatial scale.

No abstract