Arantza del Pozo

Vicomtech · Human Speech and Language Technology Department

This work introduces the design and assessment of a voice-controlled elevator system aimed at facilitating touchless interaction between users and hardware, thereby minimising contact and improving accessibility for individuals with disabilities. The research distinguishes three distinct deployment scenarios – on cloud, on edge and embedded – with...

Conversational speech interface development, maintenance and evolution is challenging for non-experts as it requires linguistic knowledge and proficiency in chatbot design and implementation. To address this issue, this work proposes the use of Dialogue Templates, compact conversational interfaces intended to cater specific interaction capabilities...

The population in the world is aging dramatically, and therefore, the economic and social effort required to maintain the quality of life is being increased. Assistive technologies are progressively expanding and present great opportunities; however, given the sensitivity of health issues and the vulnerability of older adults, some considerations n...

This paper addresses the problem of accomplishing Orofacial Rehabilitation (OR) with the assistance of artificial intelligence. The main challenges involve accurately monitoring and interacting with the trainees, while preserving user experience. We analyse different approaches to solving these challenges and propose a methodology to build smart kn...

Data scarcity is a common issue in the development of Dialogue Systems from scratch, where it is difficult to find dialogue data. This scenario is more likely to happen when the system’s language differs from English. This paper proposes a first text augmentation approach that selects samples similar to annotated user utterances from existing corpo...

RESIVOZ is a spoken dialogue system aimed at helping geriatric nurses easily register resident caring information. Compared to the traditional use of computers installed at specific control points for information recording, RESIVOZ's hands-free and mobile nature allows nurses to enter their activities in a natural way, when and where needed. Beside...

When Dialogue Systems (DS) face real usage, a challenge to solve is managing unforeseen situations without breaking the coherence of the dialogue. One way to achieve this is by redirecting the interaction to known dialogue states in a transparent way. This work proposes a simple a-priori pruning method to rule out invalid candidates when searching...

Designing dialogue policies that take user behavior into account is complicated due to user variability and behavioral uncertainty. Attributed probabilistic finite-state bi-automata (A-PFSBA) have proven to be a promising framework to develop dialogue managers that capture the users’ actions in its structure and adapt to them online, yet developing...

This paper describes the participation of Vicomtech's team in the MEDDOCAN: Medical Document Anonymization challenge, which consisted in the recognition and classification of protected health information (PHI) in medical documents in Spanish. We tested different state-of-the-art classification algorithms, both deep and shallow, and rich sets of fea...

Power point presentation of the paper "Goal-Conditioned User Modeling for Dialogue Systems using Stochastic Bi-Automata"

User simulation is widely used to generate artificial dialogues in order to train statistical spoken dialogue systems and perform evaluations. This paper presents a neural network approach for user modeling that exploits an encoder-decoder bidirectional architecture with a regularization layer for each dialogue act. In order to minimize the impact...

User Models (UM) are commonly employed to train and evaluate dialogue systems as they generate dialogue samples that simulate end-user behavior. This paper presents a stochastic approach for user modeling based in Attributed Probabilistic Finite State Bi-Automata (A-PFSBA). This framework allows the user model to be conditioned by the dialogue goal...

Data privacy compliance has gained a lot of attention over thelast years. The automation of the de-identification process is a challengingtask that often requires annotating in-domain data from scratch, as thereis usually a lack of annotated resources for such scenarios. In this work,knowledge from a classifier learnt from a source annotated datase...

In this paper the ES-Port corpus is presented. ES-Port is a spontaneous spoken human-human dialogue corpus in Spanish that consists of 1170 dialogues from calls to the technical support department of a telecommunications provider. This paper describes its compilation process, from the transcription of the raw audio to the anonymisation of the sensi...

User simulation is widely used to generate artificial dialogues in order to train statistical spoken dialogue systems and perform evaluations. This paper presents a neural network approach for user modeling that exploits an encoder-decoder bidirectional architecture with a regularization layer for each dialogue act. In order to minimize the impact...

Online learning of dialogue managers is a desirable but often costly property to obtain. Probabilistic Finite State Bi-Automata (PFSBA) have shown to provide a flexible and adaptive framework to achieve this goal. In this paper, an Attributed PFSBA (A-PSFBA) is implemented and experimentally compared with previous non-attributed PFSBA proposals. Th...

Automatic segmentation of subtitles is a novel research field which has not been studied extensively to date. However, quality automatic subtitling is a real need for broadcasters which seek for automatic solutions given the demanding European audiovisual legislation. In this article, a method based on Conditional Random Field is presented to deal...

A frequent difficulty faced by developers of Dialog Systems is the absence of a corpus of conversations to model the dialog statistically. Even when such a corpus is available, neither an agenda nor a statistically-based dialog control logic are options if the domain knowledge is broad. This article presents a module that automatically generates sy...

This paper describes the evaluation methodology followed to measure the impact of using a machine learning algorithm to automatically segment intralingual subtitles. The segmentation quality, productivity and self-reported post-editing effort achieved with such approach are shown to improve those obtained by the technique based in counting characte...

This paper describes the evaluation methodology followed to measure the impact of using a machine learning algorithm to automatically segment intralingual subtitles. The segmentation quality, productivity and self-reported post-editing effort achieved with such approach are shown to improve those obtained by the technique based in counting characte...

Technology roadmapping provides a strategic tool to help companies develop an outside-in view and challenge their current competitive perspectives. In this paper, the authors describe the roadmapping process, which is aligned, with the research and development (R&D) strategy of an applied research centre. This process is based in an adapted combina...

Using dialog systems to automatize customer services is becoming a common practice in many business fields. These dialog systems are often required to relate the users’ issues with a department of the company, which is especially hard when each department covers a wide range of topics. This paper proposes an entropy-based classifier to support the...

The subtitling demand of multimedia content has grown quickly over the last years, especially after the adoption of the new European audiovisual legislation, which forces to make multimedia content accessible to all. As a result, TV channels have been moved to produce subtitles for a high percentage of their broadcast content. Consequently, the mar...

The demand for Access Services has quickly grown over the years, mainly due to National and International laws. This trend is expected to consolidate for subtitling in particular, as almost every broadcaster is nowadays working with digital content: large amounts of existing assets are going to be digitized in the near future. In terms of accessibi...

This paper describes the data collection, annotation and sharing activities carried out within the FP7 EU-funded SAVAS project. The project aims to collect, share and reuse audiovisual language resources from broadcasters and subtitling companies to develop large vocabulary continuous speech recognisers in specific domains and new languages, with t...

The subtitling demand has grown quickly over the years. The path of manual subtitling is no longer feasible, due to increased costs and reduced production times. Assisted Subtitling is an emerging technique, consisting in the application of Automatic Speech Recognition (ASR) to automatically generate program transcripts. This paper will report on r...

The Basque language is both a minority language (only a small proportion of the population of the Basque Country speaks it) and also a less-resourced language (being spoken only in a small region by few speakers). Fortunately, the Basque regional government is committed to its recovery, and has adopted policies for funding, among other things, lang...

This paper describes the data collection and parallel corpus compilation activities carried out in the FP7 EU-funded SUMAT project. This project aims to develop an online subtitle translation service for nine European languages combined into 14 different language pairs. This data provides bilingual and monolingual training data for statistical mach...

This position paper presents the authors' goals on advanced human computer interaction and 3D Web. Previous work on speech, natural language processing and visual technologies has achieved the development of the BerbaTek language learning demonstrator, a 3D virtual tutor that supports Basque language students through spoken interaction. Next steps...

The Basque language is one of the oldest alive in Europe, although it has suffered continuous regression over the last centuries. However, many citizens and local or regional governments have been promoting its recovery since the 1970s. Now Basque holds partial co-official language status in the Basque regions of Spain but it has no official standi...

Automatic subtitling of television content has become an approachable challenge due to the advancement of the technology involved. In addition, it has also become a priority need for many Spanish TV broadcasters, who will have to broadcast up to 90% of subtitled content by 2013 to comply with recently approved national audiovisual policies. APyCA,...

AnHitz is a prototype of a virtual Basque-speaking 3D expert that can answer questions or perform cross-lingual searches on science and technology, and show the search results in Basque by means of machine translation. It has been named after the 3-year strategic research project on language, speech and visual technologies for Basque carried out by...

The aim of the AnHitz project, whose participants are research groups with very different backgrounds, is to carry out research on language, speech and visual technologies for Basque. Several resources, tools and applications have been developed in AnHitz, but we have also integrated many of these into a prototype of a 3D virtual expert on science...

Anhitz es un prototipo representado por un personaje virtual capaz de dar respuesta a preguntas relacionadas con la ciencia y la tecnología, integrando para tal propόsito múltiples tecnologías lingüísticas.

Most Voice Conversion (VC) systems exploit source-filter de- composition based on linear prediction (LP) to transform spec- tral envelopes, incurring as a result various issues related to the oversimplification of the LP voice source model. Whilst resid- ual prediction methods can mitigate this problem, they cannot be used to modify voice source qu...

This paper describes an investigation into the repair of the prosodic limitations of tracheoesophageal (TE) speech. The proposed repair algorithm modifies TE phone durations based on the predictions of regression trees built from non- pathological data. Acoustic and language modelling refinements for improved TE phone recognition, studies of featur...

This paper describes an investigation into the repair of con- tinuous tracheoesophageal (TE) speech. Our repair system resynthesises TE speech using a synthetic glottal waveform, reduces its jitter and shimmer and applies a novel spectral smoothing and tilt correction algorithm, derived from a comparative study of normal and TE spectral envelopes....