API for IT and Webmasters

API’s available from Tywi:

For existing Customer applications, Tywi provides APIs that can be integrated and used to automatically localize existing applications. Tywi APIs include:

  • Speech recognition (speech-to-text) that creates captions
  • Text-to text translation of the caption text with special dictionaries plus Tywi’s Context Engine
  • Text-to-speech (TTS) synthesized audio to speak the translated text (voice-to-voice)
  • Pre-translated text by professional translators (lookup)
  • Click to summon interpreter on demand
  • Simultaneous Interpretation software for real time-human translation (style United Nations)

Apps available

Tywi Medical app: A usable, customizable app for multi-language communication is available on mobile or any other device at a click.   Can be used by any facility, enterprise, hospital, or medical office, as well as for telemedicine. Includes ability for non-English speakers, the Deaf, the Blind, and the Deaf-Blind to make appointments for medical care by communicating directly with the reception or contact center staff. Additionally, it enables medical staff to speak with patients inside the facility; urgent care and caregivers can communicate with their clients; and large fonts are displayed for the visually impaired and aged. When 100% perfect translation is required, the “pre-translation” feature feeds perfect translations by professional translators from a database into the app fields and speaks the perfect translation at a touch.

Tywi E-911 app: E-911 is designed for communication between persons involved in an emergency and First Responders such as police, fire fighter, and ambulance. The app includes across-language text and voice translation with option to add GPS locator. A click permits first responders to communicate with the victim and family. Another click adds a medical or signing human interpreter. Medical dictionaries and personalized dictionaries are available and integrate instantly into the app.

Tywi Web Conferencing: Includes full capabilities for web conferencing, then adds multi-language translated conversation, screen share, camera, international text chat, instant captions and subtitles, plus both-way speech-to-speech translation.

 

Tywi Basic Components

Tywi aggregates the best state-of-the-art realtime language automation services in the industry, along with proprietary inventions and AI, and places them all at the fingertips of developers, including:

  • 78 text languages
  • 45 voice languages
  • 6+ automatic translation engines
  • 5+ speech recognition engines
  • Tywi Context Engine
  • Professional dictionaries
  • Personal dictionaries
  • High security web microphone plus double encryption
  • Collaboration tools (screen share, dial-in telephone, camera, file presentation, and more)

Plus Tywi provides improvements to automatic translation results and a “learning” capability.  And Tywi’s streaming professional interpretation software (human interpretation United Nations’ style) can be added to any application, the first of its kind in the world.

The Tywi APIs use streaming RTC (Real-Time-Communication) that transfers information from browser to browser (double encrypted) with only milliseconds between steps.  All data is transient, no content is stored on Tywi servers (no conversations, text or audio).   Developers can create their own web conferencing, insert Tywi in their website, or use any of the Tywi APIs in their own applications. Upon request, other types of data transfer can be provided.

Tywi Secure Web Microphone

Tywi offers a powerful invention: the Tywi Secure Web Microphone.  The Tywi “soft microphone” is installed directly on the webpage.  Where other applications stream spoken audio to their servers in a manner vulnerable to hacking, and then break that audio into smaller chunks on their servers to send to speech recognition, the Tywi Secure Web Microphone performs all chunking on the client side in a manner that creates the highest security in audio transfer on the Internet, at the same time as it meets HIPAA standards, and saves up to 60% of server space. More later in this document.

Speech Recognition (Speech–to–Text)

Send audio and receive the transcribed text results in the form of subtitles or transcripts.

Designed for use with the Tywi Secure Web Microphone (although can be used with other options), the Tywi Speech Recognition API offers a variety of speech recognition options to transcribe what the speaker says (speech-to-text) with good accuracy, and accuracy is not as easy as it sounds.  Some speech recognition engines are better than others for specific languages or dialects or voices, some train more easily, some are cheaper, etc.

For example, one Portuguese speech recognition engine may recognize what someone says in Sao Paolo, whereas a different speech recognition is required to understand someone in Rio de Janeiro. With the Tywi API, developers and users select which speech recognition to use, or include them all for optimal success with a global application.

For some speech recognition engines, a “Voice Profile” feature (basic and advanced) can be added, logged and associated with a speaking voice.  A voice profile is a basic training of 1 minute or an advanced training of 10 minutes increases the software’s comprehension of a speaker’s words by up to 10% and learns every time it is used.  The concept of voice profiles is new to the public.  However soon, voice profiles will be as common as online “User Profiles” that gather information in text form.

 

Text–to–Text Translation

Send text for translation and instantly receive translated text for blogs, bots, and applications.

Whether the text comes from a website, chat text, text string, or a translation of results from speech recognition (above), the Tywi API offers translation into 105 languages and dialects through 6 different translation engines and 3 optional engines.   Some translation engines are better than others for certain languages or subject matters, and some have existing special dictionaries.  All are real-time. Ask a Tywi pro for a free consultation.

Tywi offers 2 other text-related features that are useful for developers:

  • Popular Phases offers the ability to have perfect translation. Translators (human) can pretranslate an unlimited numbers of phrases, sentences, and paragraphs, then the phrases are retrieved at runtime in any of 3 ways: (1) manual selection with finger or mouse, (2) typing a keyword then selecting from shortlist, or (3) Speak the sentence then Tywi searches the database and displays-plays the stored translation. This is usually associated with use in medical offices, customer service, tech support and similar.
  • Subtitle and TTS Sequence Controller. Once subtitles are created using the speech recognition API above, then the subtitles need to be controlled to appear in sequential order. Without a controller, the short text or short audio will arrived before longer text or audio. This Controller will resolve the issue.

 

Tywi Context Engine

Improving text translation results

The Tywi invention, the Tywi Context Engine, fixes and improves the results coming from automatic translation (MT) by helping to control the translation AI (artificial intelligence) to improve grammar issues. Many of the issues do not exist in the English.

The Tywi Context Engine helps direct translation software into the desired mode of address (formal or informal), and helps differentiate between double meanings (e.g. “coat” as a jacket or “coat” as a “coat of paint”). Also rectifies disreapancies related to male versus female speakers. All translation software only return masculine versions for sentences like “I am married” or “I am tired”, whereas Tywi returns the proper female translation. Also, the Tywi Context Engine differentiates between singular “you” and plural “you”, and even between “plural-you female” and “plural-you masculine or mixed”.

 

Tywi Context Engine

Improving text translation results

Tywi enables the text to pass through professional dictionaries and personal dictionaries for best localization results. And Translate Your World retains top professional translators around the world to create personal dictionaries for any customer including medical, financial, and technical.

Users of this API can create Personal Dictionaries themselves or request professional dictionaries from Tywi. The purpose of these Personal Dictionaries is to assure that translations are exactly to customer’s own requirements. So, for example, Tywi assures that UPS will be United Parcel Service in every language, rather than auto-translated to “Servicios de Paquetes Unidos”.   And a law firm can assure that “execute” is not translated as “kill” simply by specifying the exact appropriate translation.

So, the Personal Dictionary is basically a word-swap, although Tywi also adds “Do Not Translate” and other possibilities. Again, Tywi will often perform more improvements in real-time. Personal Dictionaries will apply to any translation software you select, and creating your own dictionary is a simple list in an Excel file where Column A may be, for example, English, Column B is Spanish, Column C is Chinese, and so on.

Lastly, there are certain translation software to which Tywi connects that also have further dictionaries and even the ability to import your own documents to improve translations. Ask Tywi staff for more information.

 

Text-to-Speech (computer voices)

Send text and receive synthesized voice results as mp3 or wav.

Text-to-speech (“TTS”) is available through the Tywi API in 45 languages.  TTS is used as the last step in voice-to-voice translation, and is the first step for the visually impaired to read text or hear subtitles in a conference.   More good news about TTS is that the international computer voices available through this API are quite attractive.  Gone are the days of choppy, stilted speech.  This text-to-speech is so nice that is used as narration for Tywi’s video translation feature.

The Subtitle and TTS Sequence Controller is available and can be used to assure that the TTS audio plays in sequential order. Without a controller, the short audio will arrive and play before longer audio.

 

Web Conferencing and Collaboration Tools

Add web tools for sharing screen, camera, files, audio, video, and telephone connection.

The Tywi API components can be used with almost any other system that accesses the internet.

If you are looking for collaboration tools such as screen share, camera share, file presentation, and similar in order to make a global communication system, you may want to try the Voxeet collaboration suite.   Voxeet is extremely compatible with Tywi, easy to use, and includes inventions such as HD quality audio at low bandwidth.  For more information go:  http://www.voxeet.com/api-sdk

For API key and credentials, contact: support@translateyourworld.com

 

Compatibility

Tywi is a Cloud Web API and we can be accessed from almost all devices and languages. A short list:

IOS (APPS)
.NET
C#
Javascript (AJAX)
Jquery
C (LibCurl)
Go
NodeJS
PHP
Phython
Ruby
and more