Wiro - Speech to Text icon

Wiro - Speech to Text

Whisper is a pre-trained model for automatic speech recognition (ASR) and speech translation. T

Wiro - Speech to Text

AI-generated

Overview

This node performs automatic speech recognition (ASR) and speech translation using the Whisper model via the Wiro API. It processes an input audio file URL, transcribes the speech to text, and optionally detects multiple speakers and diarization. This node is useful for converting audio content into text for transcription, captioning, or further text analysis.

Use Case Examples

  1. Transcribing a podcast episode to text for content creation.
  2. Generating subtitles for a video in multiple languages.
  3. Analyzing customer service calls to identify different speakers and extract conversation text.

Properties

Name Meaning
Input Audio URL URL of the audio file to be processed for speech-to-text conversion.
Language Language of the audio to improve transcription accuracy.
Max New Tokens Maximum number of tokens to generate in the transcription output.
Chunk Length Length in seconds of each audio chunk to be processed separately.
Batch Size Number of audio chunks processed at once in a batch.
Number Of Speakers Number of distinct speakers to detect in the audio.
Diarization Enable or disable speaker diarization to separate speech by different speakers.

Output

JSON

  • taskid - Identifier of the transcription task assigned by the Wiro API.
  • url - URL to access the completed transcription result.
  • status - Status of the transcription task, e.g., 'completed' or 'failed'.

Dependencies

  • Wiro API with API key and secret for authentication

Troubleshooting

  • If the node throws an error about missing or invalid task ID or socket access token, verify that the Wiro API credentials are correctly configured and the API service is reachable.
  • If transcription fails or status is 'failed', check the input audio URL is accessible and the audio format is supported.
  • Ensure the language parameter matches the spoken language in the audio for better accuracy.
  • If diarization or speaker detection is enabled but results are inaccurate, adjust the number of speakers parameter to match the actual number of speakers in the audio.

Links

Discussion