We look at the number of languages which you can use with your content. Remember, each languages is potentially a new market, and care needs to be taken to properly target your preferred leads.
There are two kinds of transcription, the first is Speech-To-Text Transcription, the second is a Text-To-Speech Transcription. In both cases, speech is involved, hence these are referred to as transcription AI’s. Similarly, with translation, what we are actually using is a Text-To-Text Translation AI.
You should also know, that technically transcription is the mapping of sounds of one language into another. However, in common parlance, especially when talking in the context of using AI’s, it means using the sounds to get the AI to work out the words.
Transliteration is a little bit different, and does not use an AI. We simply use a mapping to achieve the desired result, and as the mapping can be phonetic, it requires a human to make sure it is done correctly.
Transcription
From Wikipedia, transcription in the linguistic sense is the systematic representation of language in written form. The source can either be utterances (speech or sign language) or pre-existing text in another writing system.
Speech-To-Text
The following languages are available within the Speech-To-Text AI. The context here is, (i) upload video, (ii) trigger transcription, (iii) select dialect as required, and the transcription AI will do the rest.
< img src=“https://storage.googleapis.com/site-assets-prod/blog_assets/what_languages_are_available_textbox_3.jpg” alt=“Select a dialect to transcribe a French video” caption=“Select a dialect to transcribe a French video” >
Please remember to check the captions (play the video) and make corrections are required. The rule of thumb here is, use the AI to do the heavy lifting, and get a (human) subject matter expert to check the results for your subject matter’s (i.e. your industry) specific acronyms/words.
< bootstrap-table “table table-dark table-striped table-bordered” >
Afrikaans
Aramaic
Arabic
Armenian
Azerbaijani
Bulgarian
Bengali
Catalan
Chinese
Czech
Danish
Dutch
German
English
Spanish
Basque
Filipino
Finnish
French
Galician
Georgian
Greek
Gujarati
Hebrew
Hindi
Croatian
Icelandic
Indonesian
Italian
Japanese
Javanese
Kannada
Khmer
Korean
Lao
Latvian
Lithuanian
Hungarian
Malay
Malayalam
Marathi
Nepali
Norwegian
Persian
Polish
Portuguese
Romanian
Russian
Serbian
Sinhala
Slovak
Slovenian
Sundanese
Swahili
Swedish
Tamil
Telugu
Thai
Turkish
Urdu
Ukrainian
Vietnamese
Zulu
< /bootstrap-table >
- Please note, we are working on adding additional AI’s for languages not currently covered. If you require a specific language, or your firm wants to plug a custom AI into our platform, please reach out.
- Additionally, each language can have several dialects - Spanish for example has 20 dialects - please ensure you select the appropriate dialect.
Text-To-Speech
The following languages are available within the Text-To-Speech AI. The context here is, (i) add audio, (ii) add the *.srt or simply the text, (iii) trigger the transcribe.
< bootstrap-table “table table-dark table-striped table-bordered” >
Dutch
English
French
German
Italian
Japanese
Korean
Portugese
Spanish
Swedish
Turkish
< /bootstrap-table >
- Please note, additional voices are available, but use a less advanced AI (i.e. the synthetic voices feel more robotic). Please contact us for more information.
Translation
The following languages are available within the Text-To-Text AI. The context here is, (i) upload content, (ii) trigger translation, (iii) select dialect as required, and the translation AI will do the rest. Below, the Automatic is the AI, while Human 1 is a in-house or external resource.
< img src=“https://storage.googleapis.com/site-assets-prod/blog_assets/what_languages_are_available_textbox_4.jpg” alt=“Select a language to translate a French video” caption=“Select a language to translate a French video” >
Again, it should be stressed, the underlying AI is only as good as its training. Depending on the information complexity of your content (i.e. legal, medical, or just complicated) it is highly recommended you use the AI to do the heavy lifting, and get a subject matter expert (i.e. a professional translator) to check the results.
In practise, it is likely to be a far more effective use of the translators time, allowing them to focus on the high value content, and be less involved in the simple translations.
< bootstrap-table “table table-dark table-striped table-bordered” >
Afrikaans
Arabic
Bengali
Bosnian
Bulgarian
Cantonese
Catalan
Chinese
Croatian
Czech
Creole
Danish
Dutch
English
Estonian
Fijian
Filipino
Finnish
French
German
Greek
Creole
Hebrew
Hindi
Hmong Daw
Hungarian
Icelandic
Indonesian
Italian
Japanese
Kiswahili
Korean
Latvian
Malagasy
Malay
Maltese
Norwegian
Persian
Polish
Portuguese
Romanian
Russian
Samoan
Serbian
Slovak
Slovenian
Spanish
Swedish
Tahitian
Tamil
Thai
Tongan
Turkish
Urdu
Ukrainian
Vietnamese
Zulu
< /bootstrap-table >
Transliteration
Transliteration is a type of conversion of a text from one script to another that involves swapping letters in predictable ways. In the platform, this is done using a phonetic mapping.
Generally speaking, your subscription will provide capacity for all the required languages. The below is an image of a textbox with three language options, English, German and Hindi.
< img src=“https://storage.googleapis.com/site-assets-prod/blog_assets/what_languages_are_available_textbox_1.jpg” alt=“Text: What is your name?” caption=“Text: What is your name?” >
Now, we are going to do the same in Hindi, first we change the language, and then select the phonetic option (फोनेटिक), and then type out tumhaara naam kya he? in the English language keyboard to get below.
< img src=“https://storage.googleapis.com/site-assets-prod/blog_assets/what_languages_are_available_textbox_2.jpg” alt=“Text: तुमहारा नाम कया हे?” caption=“Text: तुमहारा नाम कया हे?” >
This is an implementation of the work done at the Wikimedia foundation. Please contact us if you have any questions around this implementation, or for corrections.
Conclusion
Please connect with us on LinkedIn, YouTube or Facebook for any comments, questions, or just to keep up to date with the work we do!
We are very grateful for your support!
Should you have any questions, or just want to drop us a note saying hello, please feel free to send us an email at hello@videotranslator.ai.
