So, Google decided to retire Google Wave before I even got a chance to get invited (lol), but let’s talk about another Google project that may have bigger potential usage than Wave – Google Universal Translator. The idea to actually call a foreigner and casually talk to them without understanding their language seems rather miraculous even by today’s standards.
Motivation & Functionality of the tool
Universal Translator, Google’s impracticable technology is aimed at converting spoken words into a different language. At present one can convert text in 52 languages using “Google Translate”. However, Universal Translator is a speech-to-speech automated translator that will translate speech in real-time. Universal Translator will be embedded in the Android phones.
The idea inspiring the development of this pioneering tool is to enable Android users to easily communicate with people speaking in other languages with their smartphones.
The working of Universal Translator is straightforward. The tool will “listen” to the speaker (sender) at one-end and wait for them to complete until it comprehends the full meaning of the words (or phrases). It will then send this interpreted data to Google’s servers for translation. The person on the other end (receiver) will be able to listen to a computerized voice translation of the phrases spoken by the speaker. The communication will proceed similarly, both ways.
Overview of the technology:
Universal Translator is essentially a combination of two existing technologies by Google.
- Google Translate: Online universal translator service.
- Voice recognition system: Technology to recognize human voice. This technology is embedded in the Nexus One phone by Google and is used to bring about users spoken commands.
As Universal Translator combines voice recognition with Google Translate, it is very important for voice reorganization to be on the ball. However voice recognition is still not very capable in current smartphones, causing users to repeat commands often. Combining these technologies will present problems for the engineers working on this project.
Another prime challenge the team of programmers will face is the problems associated to various accents used for the same language. We know that languages are spoken with different dialects across different locations. Recognizing and translating these accents in real-time will be challenging, and it’s one of the main kinks that need to be ironed out; but I think we all look forward to when such technology becomes reality. And I’m sure this one is going to get used a lot.