The ArtiPhon Task at Evalita 2016
نویسنده
چکیده
English. Despite the impressive results achieved by ASR technology in the last few years, state-of-the-art ASR systems can still perform poorly when training and testing conditions are different (e.g., different acoustic environments). This is usually referred to as the mismatch problem. In the ArtiPhon task at Evalita 2016 we wanted to evaluate phone recognition systems in mismatched speaking styles. While training data consisted of read speech, most of testing data consisted of single-speaker hypoand hyper-articulated speech. A second goal of the task was to investigate whether the use of speech production knowledge, in the form of measured articulatory movements, could help in building ASR systems that are more robust to the effects of the mismatch problem. Here I report the result of the only entry of the task and of baseline systems. Italiano. Nonostante i notevoli risultati ottenuti recentemente nel riconoscimento automatico del parlato (ASR) le prestazioni dei sistemi ASR peggiorano significativamente in quando le condizioni di testing sono differenti da quelle di training (per esempio quando il tipo di rumore acustico è differente). Un primo gol della ArtiPhon task ad Evalita 2016 è quello di valutare il comportamento di sistemi di riconoscimento fonetico in presenza di un mismatch in termini di registro del parlato. Mentre il parlato di training consiste di frasi lette ad un velocita; di eloquio “standard”, il parlato di testing consiste di frasi sia iperche ipo-articolate. Un secondo gol della task è quello di analizzare se e come l’utilizzo di informazione concernente la produzione del parlato migliora l’accuratezza dell’ASR e in particolare nel caso di mismatch a livello di registri del parlato. Qui riporto risultati dell’unico sistema che è stato sottomesso e di una baseline.
منابع مشابه
Phone Recognition Experiments on ArtiPhon with KALDI
English. In this work we present the results obtained so far in different recognition experiments working on the audio only part of the ArtiPhon corpus used for the EVALITA 2016 speech-mismatch ArtiPhon task. Italiano. In questo lavoro si presentano i risultati ottenuti sinora in diversi esperimenti di riconoscimento fonetico utilizzanti esclusivamente la sola parte audio del corpus ArtiPhon ut...
متن کاملMivoq Evalita 2016 PosTwITA tagger
English. The POS tagger developed by Mivoq to tag tweets according to PosTwITA task guidelines as defined at Evalita 2016 is presented. The system obtained third position with 92.7% of accuracy. Italiano. Si presenta il POS tagger sviluppato da Mivoq per etichettare i tweet secondo le linee guida del task PosTwITA, cos come definite per Evalita 2016. Il sistema ha ottenuto la terza posizione co...
متن کاملThe EVALITA 2016 Event Factuality Annotation Task (FactA)
English. This report describes the FactA (Event Factuality Annotation) Task presented at the EVALITA 2016 evaluation campaign. The task aimed at evaluating systems for the identification of the factuality profiling of events. Motivations, datasets, evaluation metrics, and postevaluation results are presented and discussed. Italiano. Questo report descrive il task di valutazione FactA (Event Fac...
متن کاملOverview of the Evalita 2016 SENTIment POLarity Classification Task
English. The SENTIment POLarity Classification Task 2016 (SENTIPOLC), is a rerun of the shared task on sentiment classification at the message level on Italian tweets proposed for the first time in 2014 for the Evalita evaluation campaign. It includes three subtasks: subjectivity classification, polarity classification, and irony detection. In 2016 SENTIPOLC has been again the most participated...
متن کاملsisinflab: an Ensemble of Supervised and Unsupervised Strategies for the NEEL-IT Challenge at Evalita 2016
English. This work presents the solution adopted by the sisinflab team to solve the task NEEL-IT (Named Entity rEcognition and Linking in Italian Tweets) at the Evalita 2016 challenge. The task consists in the annotation of each named entity mention in a Twitter message written in Italian, among characters, events, people, locations, organizations, products and things and the eventual linking w...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2016