Showing papers by "Orange S.A. published in 2005"

The 2005 PASCAL visual object classes challenge

Abstract: The PASCAL Visual Object Classes Challenge ran from February to March 2005. The goal of the challenge was to recognize objects from a number of visual object classes in realistic scenes (i.e. not pre-segmented objects). Four object classes were selected: motorbikes, bicycles, cars and people. Twelve teams entered the challenge. In this chapter we provide details of the datasets, algorithms used by the teams, evaluation criteria, and results achieved.

Methods and apparatus for organizing and presenting contact information in a mobile communication system

Abstract: Methods and apparatus for gathering, organizing and displaying contact information in a mobile communications network. In one embodiment, a method for operating a mobile communication network includes an information server and a mobile communications device coupled to the service via a wireless connection. The method comprises transmitting a location identifier message including a current location from the mobile communications device to the server via the wireless link. The mobile device also transmits a list of contacts to the server via the wireless link. The contacts are matched with the location using a location database maintained for other mobile communications devices in order to generate a list of proximate users. Alternatively, the contact list can be based upon cognitive or psychographic proximity (versus merely physical proximity). The list of proximate users is transmitted to the mobile communications device, and the list is displayed in an easily perceived fashion based on user preferences.

On automating Web services discovery

Abstract: One of the challenging problems that Web service technology faces is the ability to effectively discover services based on their capabilities. We present an approach to tackling this problem in the context of description logics (DLs). We formalize service discovery as a new instance of the problem of rewriting concepts using terminologies. We call this new instance the best covering problem. We provide a formalization of the best covering problem in the framework of DL-based ontologies and propose a hypergraph-based algorithm to effectively compute best covers of a given request. We propose a novel matchmaking algorithm that takes as input a service request (or query) Q and an ontology $\mathcal{T}$ of services and finds a set of services called a “best cover” of Q whose descriptions contain as much common information with Q as possible and as little extra information with respect to Q as possible. We have implemented the proposed discovery technique and used the developed prototype in the context of the Multilingual Knowledge Based European Electronic Marketplace (MKBEEM) project.

Active attack against HB+: a provably secure lightweight authentication protocol

Abstract: Much research has focused on providing RFID tags with lightweight cryptographic functionality. The HB+ authentication protocol was recently proposed and claimed to be secure against both passive and active attacks. A linear-time active attack against HB+ is proposed.

Content navigation service

Abstract: A system, method, and computer program product for presenting items within a list of items based on an individual user profile, a social network of friends of the user or a mood of the user. The system generates the individual profile of the user, determines passive friends of the user from the social network of the user and also determines the mood of the user. The user has a choice to receive an ordered list of items based on the individual user profile, the social network or the mood of the user.

Method and apparatus for enhancing speech recognition accuracy by using geographic data to filter a set of words

Abstract: Enhanced speech recognition accuracy is provided by using geographic data, illustratively related to the geographic location of a mobile device, to automatically select a subset of words for use with a speech recognition procedure. The subset of words is selected from an element database including words that describe elements at each of a plurality of locations. Geographic data includes at least one of global positioning system (GPS) position data, cell identity (Cell-ID) data, caller identification (Caller ID) data, place name data, or zip code data. Elements include at least one of street names, businesses, merchants, points of interest, transportation facilities, individual households, activities, and landmarks. By selecting a subset of words, the total number of words used in the speech recognition procedure is substantially reduced, thereby improving speech recognition accuracy.

The finite variant property: how to get rid of some algebraic properties

Abstract: We consider the following problem: Given a term t, a rewrite system $\cal R$, a finite set of equations E′ such that $\cal R$ is E′-convergent, compute finitely many instances of t: t1,...,tn such that, for every substitution σ, there is an index i and a substitution θ such that $t\sigma\mathord\downarrow =_{E'} t_i\theta$ (where $t\sigma\mathord\downarrow$ is the normal form of tσ w.r.t. $\to_{E'\mathord{\setminus}\mathcal R}$). The goal of this paper is to give equivalent (resp. sufficient) conditions for the finite variant property and to systematically investigate this property for equational theories, which are relevant to security protocols verification. For instance, we prove that the finite variant property holds for Abelian Groups, and a theory of modular exponentiation and does not hold for the theory ACUNh (Associativity, Commutativity, Unit, Nilpotence, homomorphism).

A branch and cut algorithm for hub location problems with single assignment

Abstract: The hub location problem with single assignment is the problem of locating hubs and assigning the terminal nodes to hubs in order to minimize the cost of hub installation and the cost of routing the traffic in the network. There may also be capacity restrictions on the amount of traffic that can transit by hubs. The aim of this paper is to investigate polyhedral properties of these problems and to develop a branch and cut algorithm based on these results.

A method and apparatus for creating, directing, storing and automatically delivering a message to an intended recipient upon arrival of a specified mobile object at a designated location

Abstract: A technique to electronically deliver a message from a sender to an intended recipient based on tracking movement of a mobile object, including obtaining a message provided by the sender, obtaining a location designated by the sender for delivery of the message, tracking a specified mobile object having a position­determining device that determines its own current position, and which transmits its then current position at preset time intervals, determining from the transmitted current position whether the specified mobile object has reached the designated location, and initiating a procedure for automatic delivery of the message electronically to the intended recipient upon the specified mobile object being determined to have reached the designated location

Method for interacting with automated information agents using conversational queries

Abstract: In order to reduce or eliminate ergonomic inefficiencies arising when an automated information agent is accessed from a device having a streamlined graphical user interface, the present invention provides a mechanism for selecting and interacting with automated information agents over a text-based messaging network by means of a conversational interface that can respond to queries in accordance with the location of the device. A server sends an electronic message identifying each of a plurality of agents to the device. The streamlined graphical user interface displays an agent selection screen showing identity information for each of the plurality of agents identified in the electronic message. The device includes an input mechanism capable of accepting a user's selection of an agent. Upon receiving a first input specifying a first selected agent, the device displays an agent interaction screen for the first selected agent. At least a portion of the agent interaction screen is capable of displaying a query entered into the input mechanism. Optionally, enabling mechanisms are provided on the server to allow human intervention in response to a query.

Voice synthesis system

Abstract: A voice synthesis system for interactive voice services comprises a voice server connected to a packet network dispensing a voice service to a user terminal by executing a service file associated with the voice service. An HTTP client in the voice server transmits a request containing a text to be synthesized during execution of the service file. The service file includes an address designating a resource in a voice synthesis server connected to the packet network and a command responsive to the audio format for commanding the transmitting of the request to the voice synthesis server. An HTTP server in the voice synthesis server transmits to the voice server an audio response including the text that has been synthesized by the voice synthesis server independently of the voice server.

Towards context-aware face recognition

Abstract: In this paper, we focus on the use of context-aware, collaborative filtering, machine-learning techniques that leverage automatically sensed and inferred contextual metadata together with computer vision analysis of image content to make accurate predictions about the human subjects depicted in cameraphone photos. We apply Sparse-Factor Analysis (SFA) to both the contextual metadata gathered in the MMM2 system and the results of PCA (Principal Components Analysis) of the photo content to achieve a 60% face recognition accuracy of people depicted in our cameraphone photos, which is 40% better than media analysis alone. In short, we use context-aware media analysis to solve the face recognition problem for cameraphone photos.

WildCAT: a generic framework for context-aware applications

Abstract: We present WildCAT, an extensible Java framework to ease the creation of context-aware applications. WildCAT provides a simple yet powerful dynamic model to represent an application's execution context. The context information can be accessed by application programmers through two complimentary interfaces: synchronous requests (pull mode) and asynchronous notifications (push mode). Internally, WildCAT is designed as a framework supporting different levels of extensions, from the simple configuration of the default generic implementation to completely new implementations tailored to specific needs. A given application can mix different implementations for different aspects of its context while only depending on WildCAT's simple and unified API.

Map-aided indoor mobile positioning system using particle filter

Abstract: A framework for positioning and tracking problems in indoor environments using particle filters has been developed and applied to a WLAN location determination technique. The use of particle filtering represents an interesting way to handle the variations of the signal strength measurements. It can also aggregate in a simple way different information like signal strength and the map of a building. This type of information is very useful to obtain correct trajectories without wall-crossings. Constraining this filter on a Voronoi diagram of the building decreases the complexity of the method and improves its stability as the particles cannot be trapped in a room. In this paper, particle filtering and its constraint on a Voronoi diagram are presented, with their performances in indoor environments by combining WLAN signal strength and fingerprinting.

A new direction for quality of service: flow-aware networking

Abstract: In this paper we present the architectural implications of flow-aware networking, or FAN, a new approach for realizing QoS guarantees. FAN performs traffic control based on user-defined flows using implicit admission control and per-flow scheduling. This allows adequate performance guarantees for streaming and elastic flows without requiring class of service distinction or relying on signalled traffic specifications. We argue that FAN is a new direction for Internet quality of service that is arguably necessary as an alternative to flawed classical architectures and sufficient to meet user performance requirements in a cost-effective way.

Memory-based opportunistic multi-user beamforming

Abstract: A scheme exploiting memory in opportunistic multiuser beamforming is proposed. The scheme builds on recent advances realized in M. Sharif and B. Hassibi, 2005, in the area of multi-user downlink preceding and scheduling based on partial transmitter channel state information (CSIT). Although the preceding and scheduling done in M. Sharif and B. Hassibi, 2005 is optimal within the set of unitary precoders, it is only so asymptotically for large number of users. Secondly, this scheme is unable to exploit potential time correlation of the channel. In this paper, we show (1) that exploiting memory in the transmitter allows to fill the gap to optimality for fixed (even low) number of users for time correlated channels, (2) how such schemes can be extended to take fairness into account in the proportional fair sense

Method and system for controlling networked wireless locks

Abstract: A system including a mobile device; a lock device having a lock identification, and configured to communicate wirelessly with the mobile device; and a server having access to a database wherein (a) a key is associated with said lock identification and (b) said lock identification is associated with at least one authorized user of said lock device, the server is configured to receive from the mobile device said lock identification and a user identification and to transmit the key associated with the lock identification when the user identification corresponds to an authorized user associated with said lock identification.

Evaluating the number of active flows in a scheduler realizing fair statistical bandwidth sharing

Abstract: Despite its well-known advantages, per-flow fair queueing has not been deployed in the Internet mainly because of the common belief that such scheduling is not scalable. The objective of the present paper is to demonstrate using trace simulations and analytical evaluations that this belief is misguided. We show that although the number of flows in progress increases with link speed, the number that needs scheduling at any moment is largely independent of this rate. The number of such active flows is a random process typically measured in hundreds even though there may be tens of thousands of flows in progress. The simulations are performed using traces from commercial and research networks with quite different traffic characteristics. Analysis is based on models for balanced fair statistical bandwidth sharing and applies properties of queue busy periods to explain the observed behaviour.

Inter-cell scheduling in wireless data networks

Abstract: Over the past few years, the design and performance of channel-aware scheduling strategies have attracted huge interest. In the present paper we examine a different notion of scheduling, namely coordination of transmissions among base stations, which has received little attention so far. The inter-cell coordination comprises two key elements: (i) interference avoidance; and (ii) load balancing. The interference avoidance involves coordinating the activity phases of interfering base stations so as to increase transmission rates. The load balancing aims at diverting traffic from heavily-loaded cells to lightly-loaded cells. We consider a dynamic scenario where users come and go over time as governed by the arrival and completion of random data transfers, and evaluate the potential capacity gains from inter-cell coordination in terms of the maximum amount of traffic that can be supported for a given spatial traffic pattern. Numerical experiments demonstrate that inter-cell scheduling may provide significant capacity gains, the relative contribution from interference avoidance vs. load balancing depending on the configuration and the degree of load imbalance in the network.

Analysis of peer-to-peer traffic on ADSL

Abstract: Peer-to-Peer (P2P) applications now generate the majority of Internet traffic, particularly for users on ADSL because of flatrate tarification. In this study, we focus on four popular P2P systems to characterize the utilization, the performance and the evolution of P2P traffic in general. We observe and compare the influence of each P2P application over the traffic, and we evaluate the evolution of these P2P systems over a year. Our analysis is based on ADSL traffic captured at TCP level on a Broadband Access Server comprising thousands of users. Thus, we characterize the P2P traffic and users, and we draw interesting results on connectivity and cooperation between peers, localization of sources, termination of connections and performance limitations. The evolution of the traffic over the year allows us to see the dynamics of the use of P2P systems. The difference between week days and week-end days informs us about the behavior of P2P users.

Study of specific absorption rate (SAR) induced in two child head models and in adult heads using mobile phones

Abstract: This paper gives a first comparison of specific absorption rate (SAR) induced in a child-sized (CS) head and an adult head using a dual-band mobile phone. In the second study, the visible human head is considered and comparison of SAR induced in a CS or child-like (CL) head and an adult head using a dual-band mobile phone is given. All the peaks of average SAR over a mass of 10 and 1 g in the head and the power budget are determined in the two comparisons using the finite-difference time-domain method. The differences between the results for adult and CS or CL heads are given at 900 and 1800 MHz. No important differences are noted for the peak SAR averaged over 10 g (SAR10 g), between the two adult head models, as well as between the two child head models. The peak SAR10 g in the brain of the CS or CL head is slightly more significant than that for the adult one.

Measurement of Light attenuation in dense fog conditions for FSO applications

Abstract: Free Space Optics (FSO) has gained considerable importance in this decade of demand for high bandwidth transmission capabilities. FSO can provide the last mile solution, but the availability and reliability issues concerned with it have acquired more attention, and a need for thorough investigations. In this work, we present our results about fog attenuation at the 950 and 850 nm wavelengths in heavy maritime fog with peak values up to 500 dB/km. For the attenuation measurement, optical wavelengths are transmitted over the same path of fog in free air to a receiver, measuring the power of every wavelength. The RF marker technology employed takes advantage of modulating every optical wavelength with an individual carrier frequency, allowing to use one optical front end for the receiver and to separate individual wavelengths by electrical signal filters. The measurement of fog attenuation at different wavelengths was performed at the France Telecom R & D test facility at La Turbie. Maritime or advection fog, which caused the light attenuation consists of water droplets of larger diameter in the order of 20 μm and can cause visibilities as low as 30 meters. The visibility was measured using a transmissiometer at 550 nm. We compare our measurement data with the commonly used light attenuation models of Kruse and Kim, and present some interesting insights. The practical measurements described try to validate the models and therefore should lead to a more accurate availability prediction for FSO links.

Downlink fluid model of CDMA networks

Abstract: Classical CDMA models mostly consider hexagonal networks. In spite of their simplicity, the hexagonal networks need complex calculations. We develop an analytical fluid model of the network, simpler than the hexagonal one. It can provide explicit formulas of capacity. As an application of this model, we indicate the densification problem of a UMTS network which means the increase of the number of base stations in a limited zone to satisfy a traffic need.

Provisioning for interdomain quality of service: the MESCAL approach

Abstract: This article presents an architecture for supporting interdomain QoS across the multi-provider global Internet. While most research to date has focused on supporting QoS within a single administrative domain, mature solutions are not yet available for the provision of QoS across multiple domains administered by different organizations. The architecture described in this article encompasses the full set of functions required in the management (service and resource), control and data planes for the provision of end-to-end QoS-based IP connectivity services. We use the concept of QoS classes and show how these can be cascaded using service level specifications (SLSs) agreed between BGP peer domains to construct a defined end-to-end QoS. We illustrate the architecture by describing a typical operational scenario.

A recursive formula for multirate systems with elastic traffic

Abstract: We present a recursive formula for evaluating the per-flow throughput in a multirate system with elastic traffic, which is the analogue of the well-known Kaufman-Roberts formula used to evaluate the blocking probability in a multirate system with circuit traffic. Such a system is representative of a link in a packet-switched network like the Internet where data flows are not blocked in case of congestion but experience lower bit rates. The maximum bit rate of each flow typically corresponds to the speed of the user access line such as the DSL access lines.

Intelligent expressions of emotions

Abstract: We propose an architecture of an embodied conversational agent that takes into account two aspects of emotions: the emotions triggered by an event (the felt emotions) and the expressed emotions (the displayed ones), which may differ in real life. In this paper, we present a formalization of emotion eliciting-events based on a model of the agent’s mental state composed of beliefs, choices, and uncertainties. This model enables to identify the emotional state of an agent at any time. We also introduce a computational model based on fuzzy logic that computes facial expressions of emotions blending. Finally, examples of facial expressions resulting from the implementation of our model are shown.

User requirements for intelligent home environments: a scenario-driven approach and empirical cross-cultural study

Abstract: This paper presents the results of an empirical cross-cultural study conducted at six different sites in five European countries in the context of the EU IST-IP project AMIGO, Ambient Intelligence for the Networked Home Environment [1]. The study employed a scenario-driven approach and used quantitative and qualitative methods to elicit feedback from the target user population on concepts for intelligent home environments. The results are clustered and transformed in prioritized design guidelines.

One microphone singing voice separation using source-adapted models

Abstract: In this paper, the problem of one microphone source separation applied to singing voice extraction is studied. A probabilistic approach based on Gaussian mixture models (GMM) of the short time spectra of two sources is used. The question of source model adaptation is investigated in order to improve separation quality. A new adaptation method consisting in a filter adaptation technique via the maximum likelihood linear regression (MLLR) is presented with an associated filter-adapted training phase.