Projective active shape models for pose-variant image analysis of quasi-planar objects: Application to facial analysis

One of the important obstacles in the image-based analysis of the human face is the 3D nature of the problem and the 2D nature of most imaging systems used for biometric applications. Due to this, accuracy is strongly influenced by the viewpoint of the images, being frontal views the most thoroughly studied. However, when fully automatic face analysis systems are designed, capturing frontal-view images cannot be guaranteed. Examples of this situation can be found in surveillance systems, car driver images or whenever there are architectural constraints that prevent from placing a camera frontal to the subject. Taking advantage of the fact that most facial features lie approximately on the same plane, we propose the use of projective geometry across different views. An active shape model constructed with frontal-view images can then be directly applied to the segmentation of pictures taken from other viewpoints. The proposed extension demonstrates being significantly more invariant than the standard approach. Validation of the method is presented in 360 images from the AV@CAR database, systematically divided into three different rotations (to both sides), as well as upper and lower views due to nodding. The presented tests are among the largest quantitative results reported to date in face segmentation under varying poses.     

Sparse CCA using a Lasso with positivity constraints

Canonical correlation analysis (CCA) describes the relationship between two sets of variables by finding linear combinations of the variables with maximal correlation. A sparse version of CCA is proposed that reduces the chance of including unimportant variables in the canonical variates and thus improves their interpretation. A version of the Lasso algorithm incorporating positivity constraints is implemented in tandem with alternating least squares (ALS), to obtain sparse canonical variates. The proposed method is demonstrated on simulation studies and a data set from market basket analysis.     

Deciding k-Colorability of P 5-Free Graphs in Polynomial Time

The problem of computing the chromatic number of a P ₅-free graph (a graph which contains no path on 5 vertices as an induced subgraph) is known to be NP-hard. However, we show that for every fixed integer k, there exists a polynomial-time algorithm determining whether or not a P ₅-free graph admits a k-coloring, and finding one, if it does.     

Discriminating between disturbance and process model mismatch in model predictive control

A novel method for discriminating faults in model predictive control is presented. The proposed method monitors the Kalman filter innovations to detect the presence of autocorrelation, which is an indication of suboptimal state estimation. The cause of the suboptimal state estimation is diagnosed by the observability of this innovations process. This task involves determining the order of the autocorrelation present in the innovations. The proposed MPC fault discrimination method is demonstrated on a SISO process and a MIMO process.     

New statistical methods for investigating submarine pockmarks

We investigate the applicability of some novel spatial analysis techniques, developed for studies of astrophysical datasets, to the analysis of spatial point data in sedimentary basins. The techniques are evaluated and compared with standard methods using two test areas that contain large numbers of submarine pockmarks developed in distributed arrays. The familiar Ripley K and Voronoi tesselation techniques are used, and the results are then compared with those obtained using more novel techniques, the correlation length and minimal spanning tree. The correlation length technique is found to identify the precise distances at which clustering occurs more accurately, making a physical interpretation more clear than is possible using the Ripley K. The minimal spanning tree is found to be powerful at identifying the space-filling nature of the pockmark distribution, and has the advantage of being immune to edge effects. The use of these two novel techniques permits more information to be extracted from the datasets, and demonstrates clear statistically significant differences between them, which are not detectable using standard techniques.     

Decentralized robust control of mechanical systems

For mechanical systems described by Euler-Lagrange equations and involving high-order interconnections, this paper proposes a decentralized control, which guarantees the uniform ultimate stability when the control objective is tracking a smooth desired trajectory. The controller design is based on some physical properties of such systems and Lyapunov design methodology. The application of the proposed controller to robot manipulators is presented, and experiment results are included to illustrate the performance of the proposed control algorithm     

Real time video frames allocation in mobile networks using cooperative pre-fetching

In this paper, we study the bandwidth allocation problem for serving video requests in a mobile real-time video player system. One of the main issues to maintain the quality of services (QoS) in mobile video playback is to ensure sufficient number of video frames available at the client side while a video is being played. However, due to the limited bandwidth of a mobile network and variable workload streaming video, this is not easy to achieve in practice. In addition, the communication link between mobile clients and a video server is subject to frequent errors and temporary disconnection. In this paper, we propose to use the notion of buffered bandwidth in addition to the network bandwidth to support real-time playback of videos at mobile clients. Based on this, we designed a bandwidth allocation scheme called Cooperative Pre-Fetching (CP) in which the amount of bandwidth to be allocated to serve a video request depends on the current buffer level of the video at the client relative to the target buffer level of the client. In determining the target buffer level, we consider the errors in communication to the client as well as the other clients who are concurrently served by the system. The buffered video frames at the clients are then used to minimize the impact of error in communications on the overall QoS of video playbacks in the whole system.

Semi-automated CT segmentation using optic flow and Fourier interpolation techniques

In radiotherapy treatment planning, tumor volumes and anatomical structures are manually contoured for dose calculation, which takes time for clinicians. This study examines the use of semi-automated segmentation of CT images. A few high curvature points are manually drawn on a CT slice. Then Fourier interpolation is used to complete the contour. Consequently, optical flow, a deformable image registration method, is used to map the original contour to other slices. This technique has been applied successfully to contour anatomical structures and tumors. The maximum difference between the mapped contours and manually drawn contours was 6 pixels, which is similar in magnitude to difference one would see in manually drawn contours by different clinicians. The technique fails when the region to contour is topologically different between two slices. A solution is recommended to manually delineate contours on a sparse subset of slices and then map in both directions to fill the remaining slices.