Discriminative spoken language understanding using word confusion networks

Current commercial dialogue systems typically use hand-crafted grammars for Spoken Language Understanding (SLU) operating on the top one or two hypotheses output by the speech recogniser. These systems are expensive to develop and they suffer from significant degradation in performance when faced with recognition errors. This paper presents a robust method for SLU based on features extracted from the full posterior distribution of recognition hypotheses encoded in the form of word confusion networks. Following [1], the system uses SVM classifiers operating on n-gram features, trained on unaligned input/output pairs. Performance is evaluated on both an off-line corpus and on-line in a live user trial. It is shown that a statistical discriminative approach to SLU operating on the full posterior ASR output distribution can substantially improve performance both in terms of accuracy and overall dialogue reward. Furthermore, additional gains can be obtained by incorporating features from the previous system output. © 2012 IEEE.

Preliminary model development for predicting strength and stiffness of cement-stabilized soils using artificial neural networks

This paper presents ongoing work on data collection and collation from a large number of laboratory cement-stabilization projects worldwide. The aim is to employ Artificial Neural Networks (ANN) to establish relationships between variables, which define the properties of cement-stabilized soils, and the two parameters determined by the Unconfined Compression Test, the Unconfined Compressive Strength (UCS), and stiffness, using E50 calculated from UCS results. Bayesian predictive neural network models are developed to predict the UCS values of cement-stabilized inorganic clays/silts, as well as sands as a function of selected soil mix variables, such as grain size distribution, water content, cement content and curing time. A model which can predict the stiffness values of cement-stabilized clays/silts is also developed and compared to the UCS model. The UCS model results emulate known trends better and provide more accurate estimates than the results from the E50 stiffness model. © 2013 American Society of Civil Engineers.

Characterization and deformation response of orthotropic fibre networks with auxetic out-of-plane behaviour

In the present paper, highly porous fibre networks made of 316L fibres, with different fibre volume fractions, are characterized in terms of network architecture, elastic constants and fracture energies. Elastic constants are measured using quasi-static mechanical and modal vibration testing, yielding local and globally averaged properties, respectively. Differences between quasi-static and dynamic elastic constants are attributed to through-thickness shear effects. Regardless of the method employed, networks show signs of material inhomogeneity at high fibre densities, in agreement with X-ray nanotomography results. Strong auxetic (or negative Poisson's ratio) behaviour is observed in the through-thickness direction, which is attributed to fibre kinking induced during processing. Measured fracture energies are compared with model predictions incorporating information about in-plane fibre orientation distribution, fibre volume fraction and single fibre work of fracture. Experimental values are broadly consistent with model predictions, based on the assumption that this energy is primarily associated with plastic deformation of individual fibres within a process zone of the same order as the inter-joint spacing. © 2013 Published by Elsevier Ltd. on behalf of Acta Materialia Inc. All rights reserved.

Human mesenchymal stem cell response to 444 ferritic stainless steel networks

The aim of this work is to improve bone-implant bonding. This can, potentially, be achieved through the use of an implant coating composed of fibre networks. It is hypothesised that such an implant can achieve strong peri-prosthetic bone anchorage, when seeded with human mesenchymal stem cells (hMSCs). The materials employed were 444 and 316L stainless steel fibre networks of the same fibre volume fraction. The present work confirms that hMSCs are able to proliferate and differentiate towards the osteogenic lineage when seeded onto the fibre networks. Cellular viability, proliferation and metabolic activity were assessed and the results suggest higher proliferation rates when hMSC are seeded onto the 444 networks as compared to 316L. Cell distribution was found uniform across the seeded surfaces with 444 showing a somewhat higher infiltration depth. Copyright © Materials Research Society 2013.