An analysis of visual adaptation and contrast perception for tone mapping

Tone Mapping is the problem of compressing the range of a High-Dynamic Range image so that it can be displayed in a Low-Dynamic Range screen, without losing or introducing novel details: The final image should produce in the observer a sensation as close as possible to the perception produced by the real-world scene. We propose a tone mapping operator with two stages. The first stage is a global method that implements visual adaptation, based on experiments on human perception, in particular we point out the importance of cone saturation. The second stage performs local contrast enhancement, based on a variational model inspired by color vision phenomenology. We evaluate this method with a metric validated by psychophysical experiments and, in terms of this metric, our method compares very well with the state of the art.

The Spanish motor industry in the first third of the 20th Century: A lost opportunity

The Spanish automobile industry had a late start. Although the country proved capable of short production runs of high-quality vehicles during the first third of the century it never managed to build up its own industry, unlike Great Britain, France, or Italy. What then, were the critical shortcomings that prevented the establishment of large Spanish motor manufacturers? Put another way, why did all of the companies set up during the first half-century fail to survive? This paper attempts to shed some light on these questions, employing a wide-ranging analysis of both internal and external factors affecting the industry. A feeble internal market, lack of resources and production factors are usually adduced as reasons, as are Spain's general economic backwardness and the role played by the public authorities. However, this paper mainly focuses on the internal factors concerning company strategy and organisation. A comparison with the Italian case helps put the traditional arguments in proper perspective and highlights those covering business strategies. Finally, we argue that a broad range of factors needs to be analysed to fully understand why Spain failed to establish a motor industry.

Explaining High Economic Growth in Small Tourism Countries with a Dynamic General Equilibrium Model

This paper shows that tourism specialisation can help to explain the observed high growth rates of small countries. For this purpose, two models of growth and trade are constructed to represent the trade relations between two countries. One of the countries is large, rich, has an own source of sustained growth and produces a tradable capital good. The other is a small poor economy, which does not have an own engine of growth and produces tradable tourism services. The poor country exports tourism services to and imports capital goods from the rich economy. In one model tourism is a luxury good, while in the other the expenditure elasticity of tourism imports is unitary. Two main results are obtained. In the long run, the tourism country overcomes decreasing returns and permanently grows because its terms of trade continuously improve. Since the tourism sector is relatively less productive than the capital good sector, tourism services become relatively scarcer and hence more expensive than the capital good. Moreover, along the transition the growth rate of the tourism economy holds well above the one of the rich country for a long time. The growth rate differential between countries is particularly high when tourism is a luxury good. In this case, there is a faster increase in the tourism demand. As a result, investment of the small economy is boosted and its terms of trade highly improve.

Estimation of dynamic latent variable models using simulated nonparametric moments

Given a model that can be simulated, conditional moments at a trial parameter value can be calculated with high accuracy by applying kernel smoothing methods to a long simulation. With such conditional moments in hand, standard method of moments techniques can be used to estimate the parameter. Since conditional moments are calculated using kernel smoothing rather than simple averaging, it is not necessary that the model be simulable subject to the conditioning information that is used to define the moment conditions. For this reason, the proposed estimator is applicable to general dynamic latent variable models. Monte Carlo results show that the estimator performs well in comparison to other estimators that have been proposed for estimation of general DLV models.

RADIC II: a fault tolerant architecture with flexible dynamic redundancy

The demand for computational power has been leading the improvement of the High Performance Computing (HPC) area, generally represented by the use of distributed systems like clusters of computers running parallel applications. In this area, fault tolerance plays an important role in order to provide high availability isolating the application from the faults effects. Performance and availability form an undissociable binomial for some kind of applications. Therefore, the fault tolerant solutions must take into consideration these two constraints when it has been designed. In this dissertation, we present a few side-effects that some fault tolerant solutions may presents when recovering a failed process. These effects may causes degradation of the system, affecting mainly the overall performance and availability. We introduce RADIC-II, a fault tolerant architecture for message passing based on RADIC (Redundant Array of Distributed Independent Fault Tolerance Controllers) architecture. RADIC-II keeps as maximum as possible the RADIC features of transparency, decentralization, flexibility and scalability, incorporating a flexible dynamic redundancy feature, allowing to mitigate or to avoid some recovery side-effects.

Estimation of dynamic latent variable models using simulated nonparametric moments

Abstract. Given a model that can be simulated, conditional moments at a trial parameter value can be calculated with high accuracy by applying kernel smoothing methods to a long simulation. With such conditional moments in hand, standard method of moments techniques can be used to estimate the parameter. Because conditional moments are calculated using kernel smoothing rather than simple averaging, it is not necessary that the model be simulable subject to the conditioning information that is used to define the moment conditions. For this reason, the proposed estimator is applicable to general dynamic latent variable models. It is shown that as the number of simulations diverges, the estimator is consistent and a higher-order expansion reveals the stochastic difference between the infeasible GMM estimator based on the same moment conditions and the simulated version. In particular, we show how to adjust standard errors to account for the simulations. Monte Carlo results show how the estimator may be applied to a range of dynamic latent variable (DLV) models, and that it performs well in comparison to several other estimators that have been proposed for DLV models.

Signaling in Dynamic Contests: Some Impossibility Results

General signaling results in dynamic Tullock contests have been missing for long. The reason is the tractability of the problems. In this paper, an uninformed contestant with valuation vx competes against an informed opponent with valuation, either high vh or low vl. We show that; (i) When the hierarchy of valuations is vh ≥ vx ≥ vl, there is no pooling. Sandbagging is too costly for the high type. (ii) When the order of valuations is vx ≥ vh ≥ vl, there is no separation if vh and vl are close. Sandbagging is cheap due to the proximity of valuations. However, if vh and vx are close, there is no pooling. First period cost of pooling is high. (iii) For valuations satisfying vh ≥ vl ≥ vx, there is no separation if vh and vl are close. Bluffing in the first period is cheap for the low valuation type. Conversely, if vx and vl are close there is no pooling. Bluffing in the first stage is too costly. JEL: C72, C73, D44, D82. KEYWORDS: Signaling, Dynamic Contests, Non-existence, Sandbag Pooling, Bluff Pooling, Separating

A behavior-based scheme using reinforcement learning for autonomous underwater vehicles

This paper presents a hybrid behavior-based scheme using reinforcement learning for high-level control of autonomous underwater vehicles (AUVs). Two main features of the presented approach are hybrid behavior coordination and semi on-line neural-Q_learning (SONQL). Hybrid behavior coordination takes advantages of robustness and modularity in the competitive approach as well as efficient trajectories in the cooperative approach. SONQL, a new continuous approach of the Q_learning algorithm with a multilayer neural network is used to learn behavior state/action mapping online. Experimental results show the feasibility of the presented approach for AUVs

Dynamic modelling of an underwater vehicle

Nessie is an Autonomous Underwater Vehicle (AUV) created by a team of students in the Heriot Watt University to compete in the Student Autonomous Underwater Competition, Europe (SAUC-E) in August 2006. The main objective of the project is to find the dynamic equation of the robot, dynamic model. With it, the behaviour of the robot will be easier to understand and movement tests will be available by computer without the need of the robot, what is a way to save time, batteries, money and the robot from water inside itself. The object of the second part in this project is setting a control system for Nessie by using the model

The Impact of Industry Collaboration on Academic Research Output: A Dynamic Panel Data Analysis

The aim of this paper is to analyse the impact of university knowledge and technology transfer activities on academic research output. Specifically, we study whether researchers with collaborative links with the private sector publish less than their peers without such links, once controlling for other sources of heterogeneity. We report findings from a longitudinal dataset on researchers from two engineering departments in the UK between 1985 until 2006. Our results indicate that researchers with industrial links publish significantly more than their peers. Academic productivity, though, is higher for low levels of industry involvement as compared to high levels.

Dynamic price competition with fixed capacities

Many revenue management (RM) industries are characterized by (a) fixed capacities in theshort term (e.g., hotel rooms, seats on an airline flight), (b) homogeneous products (e.g., twoairline flights between the same cities at similar times), and (c) customer purchasing decisionslargely influenced by price. Competition in these industries is also very high even with just twoor three direct competitors in a market. However, RM competition is not well understood andpractically all known implementations of RM software and most published models of RM donot explicitly model competition. For this reason, there has been considerable recent interestand research activity to understand RM competition. In this paper we study price competitionfor an oligopoly in a dynamic setting, where each of the sellers has a fixed number of unitsavailable for sale over a fixed number of periods. Demand is stochastic, and depending on howit evolves, sellers may change their prices at any time. This reflects the fact that firms constantly,and almost costlessly, change their prices (alternately, allocations at a price in quantity-basedRM), reacting either to updates in their estimates of market demand, competitor prices, orinventory levels. We first prove existence of a unique subgame-perfect equilibrium for a duopoly.In equilibrium, in each state sellers engage in Bertrand competition, so that the seller withthe lowest reservation value ends up selling a unit at a price that is equal to the equilibriumreservation value of the competitor. This structure hence extends the marginal-value conceptof bid-price control, used in many RM implementations, to a competitive model. In addition,we show that the seller with the lowest capacity sells all its units first. Furthermore, we extendthe results transparently to n firms and perform a number of numerical comparative staticsexploiting the uniqueness of the subgame-perfect equilibrium.

Dynamic adverse selection and debt

This paper argues that the strategic use of debt favours the revelationof information in dynamic adverse selection problems. Our argument is basedon the idea that debt is a credible commitment to end long term relationships.Consequently, debt encourages a privately informed party to disclose itsinformation at early stages of a relationship. We illustrate our pointwith the financing decision of a monopolist selling a good to a buyerwhose valuation is private information. A high level of (renegotiable)debt, by increasing the scope for liquidation, may induce the highvaluation buyer to buy early at a high price and thus increase themonopolist's expected payoff. By affecting the buyer's strategy, it mayreduce the probability of excessive liquidation. We investigate theconsequences of good durability and we examine the way debt mayalleviate the ratchet effect.

Dynamic structure function in 3-4He mixtures

Relevant features of the dynamic structure function S(q,¿) in 3-4He mixtures at zero temperature are investigated starting from known properties of the ground state. Sum rules are used to fix rigorous constraints to the different contributions to S(q,¿), coming from 3He and 4He elementary excitations, as well as to explore the role of the cross term S(3,4)(q,¿). Both the low-q (phonon-roton 4He excitations and 1p-1h 3He excitations) and high-q (deep-inelastic-scattering) ranges are discussed.

High-Momentum Response of Liquid 3He

A final-state-effects formalism suitable to analyze the high-momentum response of Fermi liquids is presented and used to study the dynamic structure function of liquid 3He. The theory, developed as a natural extension of the Gersch-Rodriguez formalism, incorporates the Fermi statistics explicitly through a new additive term which depends on the semidiagonal two-body density matrix. The use of a realistic momentum distribution, calculated using the diffusion Monte Carlo method, and the inclusion of this additive correction allows for good agreement with available deep-inelastic neutron scattering data.