Speaking while driving - preliminary results on spellings in the German speechdat-car database

نویسندگان

  • Christoph Draxler
  • Klaus Bengler
  • Cristina Olaverri-Monreal
چکیده

Voice-operated devices are of particular interest in mobile environments, e.g. vehicles. They promise a natural and intuitive interface to devices and services, and they offer hands-free operation, a legal prerequisite for in-car usage in many European countries. Spelling is a common task for the operation of voice operated devices, especially under unfavorable communication conditions. This paper presents a first analysis of the error and fluency rate for 4502 utterances from the German SpeechDat-Car database. The error rate was found to be between 1.7% and 4.4% for the spelling of natural items, and between 3.6% and 7.9% for artificial letter sequences. Only 3.6% of the utterances contained hesitations. These results suggest that spelling while driving might be a suitable means of fallback interaction if specific error recovery mechanisms are implemented.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

First experiences of the German speechdat-car database collection in mobile environments

In SpeechDat-Car, speech databases for speech driven devices and services for mobile environments are collected for nine European languages. The German SpeechDat-Car installation was the first fully equipped platform within the project. It has served as a testbed for the recording software for the entire project, and as an opportunity to perform technical and organizational feasibility tests fo...

متن کامل

The speechdat-car multilingual speech databases for in-car applications: some first validation results

The main objective of SpeechDat-Car is to develop a set of speech databases to support training and testing of multilingual speech recognition applications in the car environment. SpeechDat-Car started in April 1998 in the 4th EC framework under project code LE4-8334. The duration of the project is 30 months. Equivalent and similar resources for nine languages will be created: Danish, English, ...

متن کامل

Orientel: recording telephone speech of turkish speakers in Germany

OrienTel is a project to create telephone speech databases for both the local and the business languages of the Mediterranean and the Arab Emirates. In Germany, 300 Turkish speakers speaking German were to be recorded. The database is an extension of the SpeechDat databases. This paper outlines the recording setup, the recruitment strategy and the annotation procedure. Recruiting the speakers w...

متن کامل

SpeechDat-Car Fixed Platform

SpeechDat-Car aims to develop a set of speech databases to support training and testing of multilingual speech recognition applications in the car environment. Two types of recordings compose the database. The first type consist of wideband audio signals recorded directly in the car while the second type is composed by GSM signals transmitted from the car and recorded simultaneously in a far-en...

متن کامل

Recognition Experiments with the Speechdat-car Aurora Spanish Database Using 8khz- and 16khz-sampled Signals

Like the other SpeechDat-Car databases, the Spanish one has been collected using a 16kHz sampling frequency, and several microphone positions and environmental noises. In this work, we aim at clarify whether there is any advantage in terms of recognition performance by processing the 16kHz-sampled signals instead of the usual 8kHz-sampled ones. Recognition tests have been carried out within the...

متن کامل

ذخیره 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 2001