Continuous Time Matching Constraints for Image Streams

Corresponding image points of a rigid object in a discrete sequence of images fulfil the so-called multilinear constraint. In this paper the continuous time analogue of this constraint, for a continuous stream of images, is introduced and studied. The constraint links the Taylor series expansion of the motion of the image points with the Taylor series expansion of the relative motion and orientation between the object and the camera. The analysis is done both for calibrated and uncalibrated cameras. Two simplifications are also presented for the uncalibrated camera case. One simplification is made using an affine reduction and the so-called kinetic depths. The second simplification is based upon a projective reduction with respect to the image of a planar configuration. The analysis shows that the constraint involving second-order derivatives are needed to determine camera motion. Experiments with real and simulated data are also presented.     

Mask estimation and imputation methods for missing data speech recognition in a multisource reverberant environment

We present an automatic speech recognition system that uses a missing data approach to compensate for challenging environmental noise containing both additive and convolutive components. The unreliable and noise-corrupted (“missing”) components are identified using a Gaussian mixture model (GMM) classifier based on a diverse range of acoustic features. To perform speech recognition using the partially observed data, the missing components are substituted with clean speech estimates computed using both sparse imputation and cluster-based GMM imputation. Compared to two reference mask estimation techniques based on interaural level and time difference-pairs, the proposed missing data approach significantly improved the keyword accuracy rates in all signal-to-noise ratio conditions when evaluated on the CHiME reverberant multisource environment corpus. Of the imputation methods, cluster-based imputation was found to outperform sparse imputation. The highest keyword accuracy was achieved when the system was trained on imputed data, which made it more robust to possible imputation errors.     

Strategies for Heading Off is Project Failure

Although investment in information technology and information systems continues to increase, projects continue to fail. As a result, IS projects, particularly software projects, are perceived as high risk. By categorizing types of risk, this article helps IS professionals and all project sponsors to identify classes of risk and choose the appropriate managerial behavior to mitigate each of them.     

Integrated harvesting of energy wood and pulpwood in first thinnings using the two-pile cutting method

In order to increase the harvesting volumes of energy wood and pulpwood from first thinnings, harvesting costs have to be reduced significantly. Metsäteho Oy studied the integrated harvesting of pulpwood and energy wood based on a two-pile method, where industrial roundwood (pulpwood) and energy wood fractions are stacked into two separate piles when cutting a first-thinning stand. The productivity and cost levels of the integrated, two-pile cutting method were determined, and the harvesting costs of the two-pile method were compared with those of conventional separate wood harvesting methods.In the time studies, when the size of removal was 50 dm³, the productivity in conventional whole-tree cutting was 6% higher than in integrated cutting. With a stem size of 100 dm³, the productivity of whole-tree cutting was 7% higher than in integrated cutting. The results indicated, however, that integrated harvesting based on the two-pile cutting method enables harvesting costs to be decreased to below the current cost level of separate pulpwood harvesting in first-thinning stands. The greatest cost-saving potential lies in small-sized (d_1.3 = 7-11 cm) first thinnings. The costs of forest haulage after integrated pulpwood and energy wood cutting were higher than those of separate wood harvesting because of lower removals in integrated harvesting. The results showed that when integrated wood harvesting is based on the two-pile cutting method, the removals of both energy wood and pulpwood should be more than 20-25 m³ ha⁻¹ at the integrated harvesting sites in order to achieve economically viable integrated procurement.     

Automatic geometric reasoning in structure and motion estimation

We present a system for doing automatic surveying or structure and motion analysis given 1D images of a 2D surrounding. Nothing is known about the structure of the scene features or of the motion of the camera. The system automatically identifies and tracks the image of new points and solves the structure and motion problem. One key feature of the system is the ability to hypothesize, test and incorporate simple constraints on the scene, e.g. that two object points are the same, that several points are coplanar. In this paper we develop and test the theory for automatic geometric reasoning. Ideas on hypothesis generation and testing are presented. It is also shown how to update the uncertainty representation of the database.