Learning and Exploiting Camera Geometry for Computer Vision

This work explores the use of statistical methods in describing and estimating camera poses, as well as the information feedback loop between camera pose and object detection. Surging development in robotics and computer vision has pushed the need for algorithms that infer, understand, and utilize information about the position and orientation of the sensor platforms when observing and/or interacting with their environment.

The first contribution of this thesis is the development of a set of statistical tools for representing and estimating the uncertainty in object poses. A distribution for representing the joint uncertainty over multiple object positions and orientations is described, called the mirrored normal-Bingham distribution. This distribution generalizes both the normal distribution in Euclidean space, and the Bingham distribution on the unit hypersphere. It is shown to inherit many of the convenient properties of these special cases: it is the maximum-entropy distribution with fixed second moment, and there is a generalized Laplace approximation whose result is the mirrored normal-Bingham distribution. This distribution and approximation method are demonstrated by deriving the analytical approximation to the wrapped-normal distribution. Further, it is shown how these tools can be used to represent the uncertainty in the result of a bundle adjustment problem.

Another application of these methods is illustrated as part of a novel camera pose estimation algorithm based on object detections. The autocalibration task is formulated as a bundle adjustment problem using prior distributions over the 3D points to enforce the objects' structure and their relationship with the scene geometry. This framework is very flexible and enables the use of off-the-shelf computational tools to solve specialized autocalibration problems. Its performance is evaluated using a pedestrian detector to provide head and foot location observations, and it proves much faster and potentially more accurate than existing methods.

Finally, the information feedback loop between object detection and camera pose estimation is closed by utilizing camera pose information to improve object detection in scenarios with significant perspective warping. Methods are presented that allow the inverse perspective mapping traditionally applied to images to be applied instead to features computed from those images. For the special case of HOG-like features, which are used by many modern object detection systems, these methods are shown to provide substantial performance benefits over unadapted detectors while achieving real-time frame rates, orders of magnitude faster than comparable image warping methods.

The statistical tools and algorithms presented here are especially promising for mobile cameras, providing the ability to autocalibrate and adapt to the camera pose in real time. In addition, these methods have wide-ranging potential applications in diverse areas of computer vision, robotics, and imaging.

Putting Adequate Financial and Human Resources in Healthcare for Effective Universal Health Coverage in Kenya: Lessons and Experiences from China

In recent years, most low and middle-income countries, have adopted different approaches to universal health coverage (UHC), to ensure equity and financial risk protection in accessing essential healthcare services. UHC-related policies and delivery strategies are largely based on existing healthcare systems, a result of gradual development (based on local factors and priorities). Most countries have emphasized on health financing, and human resources for health (HRH) reform policies, based on good practices of several healthcare plans to deliver UHC for their population.

Health financing and labor market frameworks were used, to understand health financing, HRH dynamics, and to analyze key health policies implemented over the past decade in Kenya’s effort to achieve UHC. Through the understanding, policy options are proposed to Kenya; analyzing, and generating lessons from health financing, and HRH reforms experiences in China. Data was collected using mixed methods approach, utilizing both quantitative (documents and literature review), and qualitative (in-depth interviews) data collection techniques.

The problems in Kenya are substantial: high levels of out-of-pocket health expenditure, slow progress in expanding health insurance among informal sector workers, inefficiencies in pulling of health are revenues, inadequate deployed HRH, maldistribution of HRH, and inadequate quality measures in training health worker. The government has identified the critical role of strengthening primary health care and the National Hospital Insurance Fund (NHIF) in Kenya’s move towards UHC. Strengthening primary health care requires; re-defining the role of hospitals, and health insurance schemes, and training, deploying and retaining primary care professionals according to the health needs of the population; concepts not emphasized in Kenya’s healthcare reforms or programs design. Kenya’s top leadership commitment is urgently needed for tougher reforms implementation, and important lessons from China’s extensive health reforms in the past decade are beneficial. Key lessons from China include health insurance expansion through rigorous research, monitoring, and evaluation, substantially increasing government health expenditure, innovative primary healthcare strengthening, designing, and implementing health policy reforms that are responsive to the population, and regional approaches to strengthening HRH.