AI-102 Question 213
Single answerYou are designing a custom speech solution for a multinational corporation that requires transcription of customer support calls in multiple languages. The solution must allow fine-tuning of the speech recognition model to account for domain-specific terminology frequently used in the company's industry. Which Azure service configuration should you use to achieve this?
- A
Use Azure Cognitive Services Speech-to-Text with a custom language model.
- B
Use Azure Cognitive Services Translator with the standard translation model.
- C
Use Azure Bot Service with integrated speech recognition.
- D
Use Azure Cognitive Services Speech-to-Text with a pre-built language model.
Show answer and explanation
Correct answer: A
Explanation
The correct choice is Azure Cognitive Services Speech-to-Text with a custom language model because it supports fine-tuning for industry-specific terminology and transcription accuracy. Pre-built models or solutions like Translator and Bot Service do not meet the requirement for custom domain-specific speech recognition.
- A. Correct.
Correct: Azure Cognitive Services Speech-to-Text allows you to create and use custom language models, which can be trained with domain-specific terminology to improve transcription accuracy.
- B. Incorrect.
Incorrect: Azure Cognitive Services Translator is primarily used for translating text between languages, not for custom speech recognition or transcription improvements.
- C. Incorrect.
Incorrect: Azure Bot Service provides conversational AI capabilities but does not support direct customization of speech recognition models.
- D. Incorrect.
Incorrect: Using a pre-built language model in Azure Speech-to-Text does not allow customization for domain-specific terminology, which is a key requirement in this scenario.