Custom dictionaries
Our Custom Dictionary functionality biases the speech recognition algorithms of Amazon Transcribe and Speechmatics towards a specific set of terms, improving transcription accuracy—especially for:
- Product names
- Technical jargon
- People’s names
- Brand-specific language
This is especially useful when you know the type of vocabulary that will appear in a given video.
Creating a Custom Dictionary
Before using the Custom Dictionary feature, you’ll need to create one:
- Open the Custom Dictionaries page from the top-right dropdown menu
- Select the language for your dictionary
- Add your phrases (ideally fewer than 1,000 terms) by
- Typing them manually or pasting in bulk (from a text editor or spreadsheet)
- Importing them from a CSV file
- Extracting terms from a document
- Pasting a list of terms
- Click Submit to save your dictionary
Note: Adding too many terms to your custom dictionary can reduce transcription quality (but will not affect latency). If your dictionary contains too many phrases, you may see a warning.
The ‘Sounds like’ field is optional and can be used for instances where the word itself doesn't give enough pronunciation detail, for example, Porsche 911, .NET, or a person or product name that may be phonetically unclear or ambiguous.
CaptionHub will do its best to prompt you where pronunciation is unclear, but you may wish to look at some Custom Dictionary examples.
Add terms directly from the caption editor
You can now add words to your Custom Dictionary directly while editing original captions:
- While editing a caption (in your original captions), highlight and right-click the word you want to add.
- Click the Add to Custom dictionary option.
- Choose the target dictionary and, if needed, edit the Text and Sounds like field.
- Hit Submit — the word is now part of your dictionary and will be used for future transcriptions.
This helps refine recognition as you work, without needing to leave the editor.
Using a Custom Dictionary in transcription
Having created your custom dictionary, you then need to specify it when you create a transcription:
- Click on the Create captions button, and choose Auto-transcribe
- If using Amazon Transcribe or Speechmatics, a custom dictionary dropdown will appear
- Make sure Amazon Transcribe or Speechmatics is enabled in Team Settings > Transcription
- Select the custom dictionary you'd like to use, and specify everything else as you would usually
Set a default dictionary
To automatically apply a specific dictionary to all original captions:
- Go to Team settings > Custom dictionaries
- Click the three-dot menu next to your dictionary
- Select Mark as default
This will be applied whether:
- You're auto-transcribing from the Original Captions tab in the UI
- You're using automation templates under the Original tab
Note
- Language variants: Custom dictionaries now work across all regional variants of a language. This improves consistency and eliminates the need for multiple region-specific dictionaries.
- For example:
- A dictionary created for English (UK) will also work with English (US), English (Global), etc., as long as the ISO language code matches.
- You're auto-transcribing from the Original Captions tab in the UI
- You're using automation templates under the Original tab
This will be applied whether: