Examkin

Question

Difficulty: HardAWS Machine Learning Services

A media streaming company is building an automated subtitle generation pipeline for its video library. The pipeline must analyze the audio track of the videos, convert the spoken words into written text, distinguish between different speakers in the conversation, and provide precise timestamps for each word to align the subtitles with the video. Which AWS service should the company integrate into their pipeline to perform this speech-to-text conversion?

  1. A
    Amazon Polly
  2. Amazon TranscribeAnswer
  3. C
    Amazon Translate
  4. D
    Amazon Lex

Answer

Amazon Transcribe
Amazon Transcribe is the correct service because it uses automatic speech recognition (ASR) to convert audio inputs to written text. It specifically provides the required capabilities of identifying when different speakers speak (speaker diarization) and providing word-level timestamps, which are necessary for sync-aligning subtitles to the video playback.

Step-by-Step Solution

1
Identify the primary technical requirement of the scenario.
The requirement is to convert spoken audio from video tracks into written text, identify individual speakers, and obtain timestamps for word alignment.
This establishes the exact type of machine learning task needed (automatic speech recognition).
2
Evaluate standard AWS Machine Learning services for speech-to-text capabilities.
Amazon Transcribe converts speech to text and includes speaker diarization and timestamping features.
This matches the service capabilities to the specified project requirements.
3
Differentiate Amazon Transcribe from other speech and language services.
Amazon Polly handles text-to-speech, Amazon Translate handles text translation, and Amazon Lex is for interactive chatbots.
This confirms that the other options do not perform the primary task of converting spoken audio into a written subtitle text format.

Key Concept

Amazon Transcribe converts spoken audio into written text with support for speaker identification and word-level timestamping.
Estimated Time:1m 30s
Rate this question

Topics

AWS Machine Learning Services

More questions from this topic