The perception of assimilation in newly learned novel words.

The present study investigated the mechanisms underlying perceptual compensation for assimilation in novel words. During training, participants learned canonical versions of novel spoken words (e.g., decibot) presented in isolation. Following exposure to a second set of novel words the next day, participants carried out a phoneme monitoring task. Here, the novel words were presented with final alternations (e.g., decibop) in carrier sentences that either licensed assimilation (viable context: Our decibop behaved badly) or did not (unviable context: Our decibop does very well). Listeners had to monitor for the underlying form of the assimilated consonant (e.g., /t/ in decibop). Results showed more responses corresponding to the underlying form in viable than in unviable contexts. This viability effect was equivalent for novel words learned on the same day and on the previous day but was absent for unexposed control items. The processing difference between exposed and control novel words supports the idea that compensation for assimilation interacts with newly acquired phonological information and suggests that contextual compensation for assimilation is enhanced by lexical knowledge. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The regular polygon detector

This paper describes a robust regular polygon detector. Given image edges, we derive the a posteriori probability for a mixture of regular polygons, and thus the probability density function for the appearance of a set of regular polygons. Likely regular polygons can be isolated quickly by discretising and collapsing the search space into three dimensions. We derive a complete formulation for efficiently recovering the remaining dimensions using maximum likelihood at the locations of the most likely polygons. Results show robustness to noise, the ability to find and differentiate different shape types, and to perform real-time sign detection for driver assistance.     

On the response of ballistic soap to one-dimensional shock loading

The effects of projectile penetration into soft tissue are often studied using the tissue simulant ballistic soap. Consequently, a full understanding of the high strain-rate response of ballistic soap is desirable. Using the plate-impact technique, key shock parameters have been measured for impacts in the range 81–968 m/s, allowing derivation of the Hugoniot equation-of-state for soap in the U_S–u_P and σ_X–u_P planes. A polynomial Hugoniot relationship was found in the U_S–u_P plane, with the general form U_S = 1.96 + 2.41u_P − 0.72u_P² (ρ₀ = 1.107 g/cc). Further, embedded lateral manganin stress gauges were employed to interrogate the evolution of lateral stress within the soap. A gradient in lateral stress, whose magnitude increased incrementally with impact stress, was apparent behind the shock for σ_X >1 GPa. It is proposed that at higher values of σ_X, increased compression of hydrocarbon chains acts to increase the materials resistance to shear, a phenomenon consistent with steric hindrance.     

Using online linear classifiers to filter spam emails

The performance of two online linear classifiers—the Perceptron and Littlestone’s Winnow—is explored for two anti-spam filtering benchmark corpora—PU1 and Ling-Spam. We study the performance for varying numbers of features, along with three different feature selection methods: information gain (IG), document frequency (DF) and odds ratio. The size of the training set and the number of training iterations are also investigated for both classifiers. The experimental results show that both the Perceptron and Winnow perform much better when using IG or DF than using odds ratio. It is further demonstrated that when using IG or DF, the classifiers are insensitive to the number of features and the number of training iterations, and not greatly sensitive to the size of training set. Winnow is shown to slightly outperform the Perceptron. It is also demonstrated that both of these online classifiers perform much better than a standard Naïve Bayes method. The theoretical and implementation computational complexity of these two classifiers are very low, and they are very easily adaptively updated. They outperform most of the published results, while being significantly easier to train and adapt. The analysis and promising experimental results indicate that the Perceptron and Winnow are two very competitive classifiers for anti-spam filtering.     

Frequency Domain Frequency Shift for Optimal Filtering of Cyclostationary Signals

Parker, G., Frequency Domain Frequency Shift for Optimal Filtering of Cyclostationary Signals, Digital Signal Processing12 (2002) 561–589Optimum reconstruction of corrupted cyclostationary signals is achieved using the filter class known as the frequency shift filter. This filter requires the received signal to be shifted by the frequencies of cyclostationarity of the signal and with a frequency domain implementation it will often be best to effect the frequency shifts directly in the frequency domain. This paper introduces techniques for exactly achieving these shifts as well as providing more computationally efficient approximate solutions.     

Whose fault is it anyway? A practical illustration of human factors in process safety

Major process accidents have typically occurred not through a single failure, but through a combination of events, some of which had contributors from past actions and failures (latent or unrevealed failures). People are integral and key features of business systems; therefore systems, tools, and equipment should be designed with the potential capabilities and limitations of people in mind. This paper demonstrates the benefits of using human factors approaches to improve system safety and reliability. Practical examples from past experience are quoted and a framework for human error prediction is described. Guidance is given on the practicalities of deriving recommendations from these types of studies.     

Improving pulse sequences for 3D diffusion-ordered NMR spectroscopy: 2DJ-IDOSY

An improved pulse sequence for the 3D DOSY experiment 2DJ-DOSY, using diffusion encoding internal to the parent 2DJ spectroscopy sequence (2DJ-IDOSY), is presented. The diffusion-encoding pulses are used to enforce the desired coherence transfer pathway, reducing the minimum experimental time by at least a factor of 4, as compared to existing techniques, and approximately doubling the signal-to-noise ratio for small molecules. The new sequence is demonstrated on a simple mixture and on a complex sample with a high dynamic range (port wine). The principle of internal diffusion encoding can be applied with profit to a range of other 3D DOSY experiments.     

Thermal and related studies of some zirconia gels

The thermal properties of a number of zirconia gels, prepared with various precipitants, have been studied by thermogravimetry, differential thermal analysis and evolved gas analysis. The results obtained show that ligands derived from the various precipitants are incorporated in these gels and strongly influence their high temperature behaviour. The observed properties are interpreted on the basis of a defect structure involving O^2- vacancies generated through the removal of these ligands at higher temperatures.