Bibliographic Details
| Title: |
Measuring and analysis of speech-to-text accuracy of some automatic speech recognition services in dynamic environment conditions. |
| Authors: |
Gyulyustan, Hasan, Hristov, Hristo, Stavrev, Stefan, Enkov, Svetoslav |
| Source: |
AIP Conference Proceedings; 2024, Vol. 3063 Issue 1, p1-7, 7p |
| Subject Terms: |
Automatic speech recognition, Text files, Human-computer interaction, Market leaders, Software as a service, Error rates |
| Company/Entity: |
Google Inc. |
| Abstract: |
In recent years, there is rise in various human-computer interaction interfaces and automatic text translation services and frameworks. In this paper, we conduct an up-to-date evaluation of commercial automatic speech recognition (ASR) software services by comparing two market leaders – Apple Siri and Google speech-to-text (STT) services. For the experiment, we select three diverse categories – conversational, scientific and fictional texts. We then record an English speaker to read those texts, pass the voice recordings to the evaluated ASR services and save the output as text files. In a real-world scenario, the recordings are never clear and that is why we do another set of experiments with artificially introduced noise of the recordings. Comparing all generated output texts to the original input, we calculate an evaluation metric, called word error rate (WER), which is a well-know and widely used indicator for measuring automatic speech-to-text accuracy. After analyzing the results, we found out that some STT services perform better on certain text categories. Finally, we discuss the results and draw conclusions. [ABSTRACT FROM AUTHOR] |
|
Copyright of AIP Conference Proceedings is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Database: |
Complementary Index |