Semantic Discriminant mapping for classification and browsing of remote sensing textures and objects

We present a new approach based on Discriminant Analysis to map a high dimensional image feature space onto a subspace which has the following advantages: 1. each dimension corresponds to a semantic likelihood, 2. an efficient and simple multiclass classifier is proposed and 3. it is low dimensional. This mapping is learnt from a given set of labeled images with a class groundtruth. In the new space a classifier is naturally derived which performs as well as a linear SVM. We will show that projecting images in this new space provides a database browsing tool which is meaningful to the user. Results are presented on a remote sensing database with eight classes, made available online. The output semantic space is a low dimensional feature space which opens perspectives for other recognition tasks. © 2005 IEEE.

Classification and choice in product design

Life is full of difficult choices. Everyone has their own way of dealing with these, some effective, some not. The problem is particularly acute in engineering design because of the vast amount of information designers have to process. This paper deals with a subset of this set of problems: the subset of selecting materials and processes, and their links to the design of products. Even these, though, present many of the generic problems of choice, and the challenges in creating tools to assist the designer in making them. The key elements are those of classification, of indexing, of reaching decisions using incomplete data in many different formats, and of devising effective strategies for selection. This final element - that of selection strategies - poses particular challenges. Product design, as an example, is an intricate blend of the technical and (for want of a better word) the aesthetic. To meet these needs, a tool that allows selection by analysis, by analogy, by association and simply by 'browsing' is necessary. An example of such a tool, its successes and remaining challenges, will be described.

Photo-assisted hydrosilylation of silicon nanoparticles: Dependence of particle size on grafting chemistry

Silicon nanoparticles between 2.5 nm and 30 nm in diameter were functionalized by means of photoassisted hydrosilylation reactions in the aerosol phase with terminal alkenes of varying chain length. Using infrared spectroscopy and nuclear magnetic resonance, the chemical composition of the alkyl layer was determined for each combination of particle size and alkyl chain length. The spectroscopic techniques were used to determine that smaller particles functionalized with short chains in the aerosol phase tend to attach to the interior (β) alkenyl carbon atom, whereas particles >10 nm in diameter exhibit attachment primarily with the exterior (α) alkenyl carbon atom, regardless of chain length. © 2011 American Chemical Society.

Effect of bend orientation on life and puncture point location due to solid particle erosion of a high concentration flow in pneumatic conveyors

An investigation into predicting failure of pneumatic conveyor pipe bends due to hard solid particle impact erosion has been carried out on an industrial scale test rig. The bend puncture point locations may vary with many factors. However, bend orientation was suspected of being a main factor due to the biased particle distribution pattern of a high concentration flow. In this paper, puncture point locations have been studied with different pipe bend orientations and geometry (a solids loading ratio of 10 being used for the high concentration flow). Test results confirmed that the puncture point location is indeed most significantly influenced by the bend orientation (especially for a high concentration flow) due to the biased particle distribution and biased particle flux distribution. © 2004 Elsevier B.V. All rights reserved.

Effect of particle concentration on erosion rate of mild steel bends in a pneumatic conveyor

Particle concentration is known as a main factor that affects erosion rate of pipe bends in pneumatic conveyors. With consideration of different bend radii, the effect of particle concentration on weight loss of mild steel bends has been investigated in an industrial scale test rig. Experimental results show that there was a significant reduction of the specific erosion rate for high particle concentrations. This reduction was considered to be as a result of the shielding effect during the particle impacts. An empirical model is given. Also a theoretical study of scaling on the shielding effect, and comparisons with some existing models, are presented. It is found that the reduction in specific erosion rate (relative to particle concentration) has a stronger relationship in conveying pipelines than has been found in the erosion tester. © 2004 Elsevier B.V. All rights reserved.

Techniques to improve the shear strength of impacted bone graft: the effect of particle size and washing of the graft.

BACKGROUND: When fresh morselized graft is compacted, as in impaction bone-grafting for revision hip surgery, fat and marrow fluid is either exuded or trapped in the voids between particles. We hypothesized that the presence of incompressible fluid damps and resists compressive forces during impaction and prevents the graft particles from moving into a closer formation, thus reducing the graft strength. In addition, viscous fluid such as fat may act as an interparticle lubricant, thus reducing the interlocking of the particles. METHODS: We performed mechanical shear testing in the laboratory with use of fresh-frozen human femoral-head allografts that had been passed through different orthopaedic bone mills to produce graft of differing particle-size distributions (grading). RESULTS: After compaction of fresh graft, fat and marrow fluid continued to escape on application of normal loads. Washed graft, however, had little lubricating fluid and better contact between the particles, increasing the shear resistance. On mechanical testing, washed graft was significantly (p < 0.001) more resistant to shearing forces than fresh graft was. This feature was consistent for different bone mills that produced graft of different particle-size distributions and shear strengths. CONCLUSIONS: Removal of fat and marrow fluid from milled human allograft by washing the graft allows the production of stronger compacted graft that is more resistant to shear, which is the usual mode of failure. Further research into the optimum grading of particle sizes from bone mills is required.

Techniques to improve the shear strength of impacted bone graft: The effect of particle size and washing of the graft

Background: When fresh morselized graft is compacted, as in impaction bone-grafting for revision hip surgery, fat and marrow fluid is either exuded or trapped in the voids between particles. We hypothesized that the presence of incompressible fluid damps and resists compressive forces during impaction and prevents the graft particles from moving into a closer formation, thus reducing the graft strength. In addition, viscous fluid such as fat may act as an interparticle lubricant, thus reducing the interlocking of the particles. Methods: We performed mechanical shear testing in the laboratory with use of fresh-frozen human femoral-head allografts that had been passed through different orthopaedic bone mills to produce graft of differing particle-size distributions (grading). Results: After compaction of fresh graft, fat and marrow fluid continued to escape on application of normal loads. Washed graft, however, had little lubricating fluid and better contact between the particles, increasing the shear resistance. On mechanical testing, washed graft was significantly (p < 0.001) more resistant to shearing forces than fresh graft was. This feature was consistent for different bone mills that produced graft of different particle-size distributions and shear strengths. Conclusions: Removal of fat and marrow fluid from milled human allograft by washing the graft allows the production of stronger compacted graft that is more resistant to shear, which is the usual mode of failure. Further research into the optimum grading of particle sizes from bone mills is required. Clinical Relevance: Understanding the mechanical properties of milled human allograft is important when impaction grafting is used for mechanical support. A simple means of improving the mechanical strength of graft produced by currently available bone mills, including an intraoperative washing technique, is described.

Bayesian classification of acoustical waveforms under environmental variability

We present in this paper a new multivariate probabilistic approach to Acoustic Pulse Recognition (APR) for tangible interface applications. This model uses Principle Component Analysis (PCA) in a probabilistic framework to classify tapping pulses with a high degree of variability. It was found that this model, achieves a higher robustness to pulse variability than simpler template matching methods, specifically when allowed to train on data containing high variability. © 2011 IEEE.

Bayesian Active Learning for Classification and Preference Learning

Information theoretic active learning has been widely studied for probabilistic models. For simple regression an optimal myopic policy is easily tractable. However, for other tasks and with more complex models, such as classification with nonparametric models, the optimal solution is harder to compute. Current approaches make approximations to achieve tractability. We propose an approach that expresses information gain in terms of predictive entropies, and apply this method to the Gaussian Process Classifier (GPC). Our approach makes minimal approximations to the full information theoretic objective. Our experimental performance compares favourably to many popular active learning algorithms, and has equal or lower computational complexity. We compare well to decision theoretic approaches also, which are privy to more information and require much more computational time. Secondly, by developing further a reformulation of binary preference learning to a classification problem, we extend our algorithm to Gaussian Process preference learning.