Tree-Structured Stick Breaking Processes for Hierarchical Data

Many data are naturally modeled by an unobserved hierarchical structure. In this paper we propose a flexible nonparametric prior over unknown data hierarchies. The approach uses nested stick-breaking processes to allow for trees of unbounded width and depth, where data can live at any node and are infinitely exchangeable. One can view our model as providing infinite mixtures where the components have a dependency structure corresponding to an evolutionary diffusion down a tree. By using a stick-breaking approach, we can apply Markov chain Monte Carlo methods based on slice sampling to perform Bayesian inference and simulate from the posterior distribution on trees. We apply our method to hierarchical clustering of images and topic modeling of text data.

THEORETICAL ROAD DAMAGE DUE TO DYNAMIC TYRE FORCES OF HEAVY VEHICLES. PART 1: DYNAMIC ANALYSIS OF VEHICLES AND ROAD SURFACES.

Theory is presented for simulating the dynamic wheel forces generated by heavy road vehicles and the resulting dynamic response of road surfaces to these loads. Sample calculations are provided and the vehicle simulation is validated with data from full-scale tests. The methods are used in the accompanying paper to simulate the road damage done by a tandem-axle vehicle.

THEORETICAL ROAD DAMAGE DUE TO DYNAMIC TYRE FORCES OF HEAVY VEHICLES. PART 2: SIMULATED DAMAGE CAUSED BY A TANDEM-AXLE VEHICLE.

The literature relating to road surface failure and design is briefly reviewed and the conventional methods for assessing the road damaging effects of dynamic tire forces are examined. A new time domain technique for analyzing dynamic tire forces and four associated road damage criteria are presented. The force criteria are used to examine the road damaging characteristics of a simple tandem-axle vehicle model for a range of speed and road roughness conditions. It is concluded that for the proposed criteria, the theoretical service life of road surfaces that are prone to fatigue failure may be reduced significantly by the dynamic component of wheel forces. The damage done to approximately five per cent of the road surface area during the passage of a theoretical model vehicle at typical highway speeds may be increased by as much as four times.

Road damage due to dynamic tyre forces measured on a public road

This paper describes a series of tests conducted on a UK trunk road, in which the dynamic tyre forces generated by over 1500 heavy goods vehicles (HGVs) were measured using a load measuring mat containing 144 capacitive strip sensors. The data was used to investigate the relative road damaging potential of the various classes of vehicles, and the degree of spatial repeatability of tyre forces present in a typical highway fleet. Approximately half the vehicles tested were found to contribute to a spatially repeatable pattern of pavement loading. On average, air suspended vehicles were found to generate lower dynamic load coefficients than steel suspended vehicles. However, air suspended vehicles also generated higher mean levels of theoretical road damage (aggregate force) than steel suspended vehicles, indicating that the ranking of suspensions depends on the pavement damage criterion used.

Feasibility of damage detection using vibration data in a population of cylinders

The feasibility of vibration data to identify damage in a population of cylindrical shells is assessed. Vibration data from a population of cylinders were measured and modal analysis was employed to obtain natural frequencies and mode shapes. The mode shapes were transformed into the Coordinate Modal Assurance Criterion (COMAC). The natural frequencies and the COMAC before and after damage for a population of structures show that modal analysis is a viable route to damage identification in a population of nominally identical cylinders. Modal energies, which are defined as the integrals of the real and imaginary components of the frequency response functions over various frequency ranges, were extracted and transformed into the Coordinate Modal Energy Assurance Criterion (COMEAC). The COMEAC before and after damage show that using modal energies is a viable approach to damage identification in a population of cylinders.

Decision tree-based context clustering based on cross validation and hierarchical priors

The standard, ad-hoc stopping criteria used in decision tree-based context clustering are known to be sub-optimal and require parameters to be tuned. This paper proposes a new approach for decision tree-based context clustering based on cross validation and hierarchical priors. Combination of cross validation and hierarchical priors within decision tree-based context clustering offers better model selection and more robust parameter estimation than conventional approaches, with no tuning parameters. Experimental results on HMM-based speech synthesis show that the proposed approach achieved significant improvements in naturalness of synthesized speech over the conventional approaches. © 2011 IEEE.

Semi-supervised video segmentation using tree structured graphical models

We present a novel, implementation friendly and occlusion aware semi-supervised video segmentation algorithm using tree structured graphical models, which delivers pixel labels alongwith their uncertainty estimates. Our motivation to employ supervision is to tackle a task-specific segmentation problem where the semantic objects are pre-defined by the user. The video model we propose for this problem is based on a tree structured approximation of a patch based undirected mixture model, which includes a novel time-series and a soft label Random Forest classifier participating in a feedback mechanism. We demonstrate the efficacy of our model in cutting out foreground objects and multi-class segmentation problems in lengthy and complex road scene sequences. Our results have wide applicability, including harvesting labelled video data for training discriminative models, shape/pose/articulation learning and large scale statistical analysis to develop priors for video segmentation. © 2011 IEEE.

Unsupervised intra-lingual and cross-lingual speaker adaptation for HMM-based speech synthesis using two-pass decision tree construction

Hidden Markov model (HMM)-based speech synthesis systems possess several advantages over concatenative synthesis systems. One such advantage is the relative ease with which HMM-based systems are adapted to speakers not present in the training dataset. Speaker adaptation methods used in the field of HMM-based automatic speech recognition (ASR) are adopted for this task. In the case of unsupervised speaker adaptation, previous work has used a supplementary set of acoustic models to estimate the transcription of the adaptation data. This paper first presents an approach to the unsupervised speaker adaptation task for HMM-based speech synthesis models which avoids the need for such supplementary acoustic models. This is achieved by defining a mapping between HMM-based synthesis models and ASR-style models, via a two-pass decision tree construction process. Second, it is shown that this mapping also enables unsupervised adaptation of HMM-based speech synthesis models without the need to perform linguistic analysis of the estimated transcription of the adaptation data. Third, this paper demonstrates how this technique lends itself to the task of unsupervised cross-lingual adaptation of HMM-based speech synthesis models, and explains the advantages of such an approach. Finally, listener evaluations reveal that the proposed unsupervised adaptation methods deliver performance approaching that of supervised adaptation.

Building damage and repair due to leakage in a deep excavation

The north-south line in Amsterdam is being built underneath the historic centre of the city. Three deep stations are being constructed in deep excavations supported by diaphragm walls. During the excavation for Vijzelgracht station, leakage through the wall resulted in large settlements and damage to historic buildings, which threatened continuation of the project. The authors analysed the cause of the leakage and the damage to the buildings. With the application of robust preventative measures at two of the deep excavations it was possible to continue the project. This paper reports on the cause of the events, the damage to the buildings and the counter-measures taken. It includes lessons learned for the project and for the foundations industry.