The geography of maternal and newborn health: the state of the art.

As the deadline for the millennium development goals approaches, it has become clear that the goals linked to maternal and newborn health are the least likely to be achieved by 2015. It is therefore critical to ensure that all possible data, tools and methods are fully exploited to help address this gap. Among the methods that are under-used, mapping has always represented a powerful way to 'tell the story' of a health problem in an easily understood way. In addition to this, the advanced analytical methods and models now being embedded into Geographic Information Systems allow a more in-depth analysis of the causes behind adverse maternal and newborn health (MNH) outcomes. This paper examines the current state of the art in mapping the geography of MNH as a starting point to unleashing the potential of these under-used approaches. Using a rapid literature review and the description of the work currently in progress, this paper allows the identification of methods in use and describes a framework for methodological approaches to inform improved decision-making. The paper is aimed at health metrics and geography of health specialists, the MNH community, as well as policy-makers in developing countries and international donor agencies.

Faster and better: a machine learning approach to corner detection.

The repeatability and efficiency of a corner detector determines how likely it is to be useful in a real-world application. The repeatability is important because the same scene viewed from different positions should yield features which correspond to the same real-world 3D locations. The efficiency is important because this determines whether the detector combined with further processing can operate at frame rate. Three advances are described in this paper. First, we present a new heuristic for feature detection and, using machine learning, we derive a feature detector from this which can fully process live PAL video using less than 5 percent of the available processing time. By comparison, most other detectors cannot even operate at frame rate (Harris detector 115 percent, SIFT 195 percent). Second, we generalize the detector, allowing it to be optimized for repeatability, with little loss of efficiency. Third, we carry out a rigorous comparison of corner detectors based on the above repeatability criterion applied to 3D scenes. We show that, despite being principally constructed for speed, on these stringent tests, our heuristic detector significantly outperforms existing feature detectors. Finally, the comparison demonstrates that using machine learning produces significant improvements in repeatability, yielding a detector that is both very fast and of very high quality.