L'àmbit de recerca del grup 'VEU' és el tractament de la parla. Investiguem tecnologies que permeten l'extracció d'informació que la veu conté: reconeixement del que es diu, l'idioma o el dialecte, característiques del parlant -qui és, la seva edat, el sexe, l'estat emocional-, la direcció del so. També treballem en la caracterització general de l'àudio, per determinar quan hi ha veu i quan hi ha altres esdeveniments acústics com música o sorolls diversos. Les tecnologies de la parla possibiliten generar veu -síntesis de veu- o modificar les seves

Recent Submissions

  • Estimation of polar depletion regions by VTEC contrast and watershed enhancing 

    Monte Moreno, Enrique; Hernández Pajares, Manuel; Lyu, Haixia; Yang, Heng; Aragon-Angel, Angela (Institute of Electrical and Electronics Engineers (IEEE), 2021)
    Article
    Open Access
    This article presents a method for determining near-Pole ionization depletion regions and troughs from global navigation satellite system (GNSS) vertical total electron content (VTEC) maps. To define the regions, we use ...
  • Frequency domain analysis and filtering of business and consumer survey expectations 

    Claveria González, Oscar; Monte Moreno, Enrique; Torra Porras, Salvador (Elsevier, 2021-03-20)
    Article
    Restricted access - publisher's policy
    The main objective of this study is two-fold. First, we aim to detect the underlying existing periodicities in business and consumer survey expectations by means of spectral analysis. We use the Welch method to extract the ...
  • Nowcasting and forecasting GDP growth with machine-learning sentiment indicators 

    Claveria González, Oscar; Monte Moreno, Enrique; Torra Porras, Salvador (2021-02-17)
    External research report
    Open Access
    We apply the two-step machine-learning method proposed by Claveria et al. (2021) to generate country-specific sentiment indicators that provide estimates of year-on-year GDP growth rates. In the first step, by means of ...
  • Syntax-driven iterative expansion language models for controllable text generation 

    Casas Manzanares, Noé; Rodríguez Fonollosa, José Adrián; Ruiz Costa-Jussà, Marta (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    The dominant language modeling paradigm handles text as a sequence of discrete tokens. While that approach can capture the latent structure of the text, it is inherently constrained to sequential dynamics for text generation. ...
  • From bilingual to multilingual neural-based machine translation by incremental training 

    Escolano Peinado, Carlos; Ruiz Costa-Jussà, Marta; Rodríguez Fonollosa, José Adrián (2020-08-02)
    Article
    Open Access
    A common intermediate language representation in neural machine translation can be used to extend bilingual systems by incremental training. We propose a new architecture based on introducing an interlingual loss as an ...
  • GeBioToolkit: automatic extraction of gender-balanced multilingual corpus of Wikipedia biographies 

    Ruiz Costa-Jussà, Marta; Li Lin, Pau; España Bonet, Cristina (European Language Resources Association (ELRA), 2020)
    Conference lecture
    Open Access
    We introduce GeBioToolkit, a tool for extracting multilingual parallel corpora at sentence level, with document and gender information from Wikipedia biographies. Despite the gender inequalities present in Wikipedia, the ...
  • Converses al voltant de la intel·ligència artificial en clau catalana 

    Ruiz Costa-Jussà, Marta; Melero Nogues, Maite (2020-12)
    Article
    Open Access
    La intel·ligència artificial mou milions d’euros i ocupa una part important de les agendes polítiques i estratègiques dels governs. En aquest article reflexionem sobre aquest concepte difús i ho fem a través de tres ...
  • Enhancing word embeddings with knowledge extracted from lexical resources 

    Biesialska, Magdalena Marta; Rafieian, Bardia; Ruiz Costa-Jussà, Marta (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    In this work, we present an effective method for semantic specialization of word vector representations. To this end, we use traditional word embeddings and apply specialization methods to better capture semantic relations ...
  • Fine-tuning neural machine translation on gender-balanced datasets 

    Ruiz Costa-Jussà, Marta; de Jorge Sánchez, Adrián (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    Misrepresentation of certain communities in datasets is causing big disruptions in artificial intelligence applications. In this paper, we propose using an automatically extracted gender-balanced dataset parallel corpus ...
  • Findings of the first shared task on lifelong learning machine translation 

    Barrault, Loïc; Biesialska, Magdalena Marta; Ruiz Costa-Jussà, Marta; Bougares, Fethi; Galibert, Olivier (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    A lifelong learning system can adapt to new data without forgetting previously acquired knowledge. In this paper, we introduce the first benchmark for lifelong learning machine translation. For this purpose, we provide ...
  • Multilingual neural machine translation: case-study for Catalan, Spanish and Portuguese romance languages 

    Vergés Boncompte, Pere; Ruiz Costa-Jussà, Marta (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    In this paper, we describe the TALP-UPC participation in the WMT Similar Language Translation task between Catalan, Spanish, and Portuguese, all of them, Romance languages. We made use of different techniques to improve ...
  • The TALP-UPC system description for WMT20 news translation task: multilingual adaptation for low resource MT 

    Escolano Peinado, Carlos; Ruiz Costa-Jussà, Marta; Rodríguez Fonollosa, José Adrián (Association for Computational Linguistics, 2020)
    Conference lecture
    Open Access
    In this article, we describe the TALP-UPC participation in the WMT20 news translation shared task for Tamil-English. Given the low amount of parallel training data, we resort to adapt the task to a multilingual system to ...

View more