Annika Hämäläinen

Impact of Age in ASR for the Elderly: Preliminary Experiments in European Portuguese

Standard automatic speech recognition (ASR) systems use acoustic models typically trained with speech of young adult speakers. Ageing is known to alter speech production in ways that require ASR systems to be adapted, in particular at the level of acoustic modeling. This paper reports ASR experiments that illustrate the impact of speaker age on speech recognition performance. A large read speech corpus in European Portuguese allowed us to measure statistically significant performance differences among age groups ranging from 60- to 90-year-old speakers. An increase of 41% relative (11.9% absolute) in word error rate was observed between 60-65-year-old and 81-86-year-old speakers. This paper also reports experiments on retraining acoustic models (AMs), further illustrating the impact of ageing on ASR performance. Differentiated gains were observed depending on the age range of the adaptation data use to retrain the acoustic models.

Speech-centric Multimodal Interaction for Easy-to-access Online Services – A Personal Life Assistant for the Elderly

Abstract The PaeLife project is a European industry-academia collaboration whose goal is to provide the elderly with easy access to online services that make their life easier and encourage their continued participation in the society. To reach this goal, the project partners are developing a multimodal virtual personal life assistant (PLA) offering a wide range of services from weather information to social networking. This paper presents the multimodal architecture of the PLA, the services provided by the PLA, and the work done in the area of speech input and output modalities, which play a key role in the application.

Multilingual Speech Recognition for the Elderly: The AALFred Personal Life Assistant.

Abstract The PaeLife project is a European industry-academia collaboration in the framework of the Ambient Assisted Living Joint Programme (AAL JP), with a goal of developing a multimodal, multilingual virtual personal life assistant to help senior citizens remain active and socially integrated. Speech is one of the key interaction modalities of AALFred, the Windows application developed in the project; the application can be controlled using speech input in four European languages: French, Hungarian, Polish and Portuguese. This paper briefly presents the personal life assistant and then focuses on the speech-related achievements of the project. These include the collection, transcription and annotation of large corpora of elderly speech, the development of automatic speech recognisers optimised for elderly speakers, a speech modality component that can easily be reused in other applications, and an automatic grammar translation service that allows for fast expansion of the automatic speech recognition functionality to new languages.

The CNG Corpus of European Portuguese Children’s Speech

Speech recognisers trained with adults’ speech do not work well with children’s speech because of the inherent acoustic and linguistic differences in the speech of these two populations. To develop speech-driven applications capable of successfully recognising children’s speech, a sufficient amount of children’s speech is needed for training acoustic models from scratch or for adapting acoustic models trained with adults’ speech. However, the availability of suitable children’s speech corpora is still limited, especially in the case of less-spoken languages. This paper describes the design, collection, transcription and annotation of a 21-hour corpus of prompted European Portuguese children’s speech collected from 510 children aged 3-10. Before the development of this corpus, European Portuguese children’s speech data have not been available at all for parts of this age range.

Automatically Recognising European Portuguese Children’s Speech

This paper reports findings from an analysis of errors made by an automatic speech recogniser trained and tested with 3-10-year-old European Portuguese childrens speech. We expected and were able to identify frequent pronunciation error patterns in the childrens speech. Furthermore, we were able to correlate some of these pronunciation error patterns and automatic speech recognition errors. The findings reported in this paper are of phonetic interest but will also be useful for improving the performance of automatic speech recognisers aimed at children representing the target population of the study.