Working of artificially intelligent language translator?

I want to do a paper presentation and want to know how a speech recognisation and translation system works. And also guide me in writing the abstract and the paper.Working of artificially intelligent language translator?
Machine translation is still at a pretty basic state. If you want to have fun some time, try Babelfish or other translation site and translate something from one language to another and then back. Or for real giggles, translate to a third language before returning to the first.



Translation is a very, very difficult problem.



It is also a very large topic. You are going to have to select some more specific piece of it to report on.



Speech recognition has come along better than translation. One interesting aspect of speech recognition that you might consider for your report is dynamic time warping, a method used to stretch or shrink utterances to match patterns for recognition. Two people could say the same sentence - or even the same word - and one might say it much more slowly than the other. How do you take those two different inputs and match them both to the same target pattern? That's where dynamic time warping comes in (and "dynamic time warping" just sounds cool, too).



Whether you report on speech recognition or translation, I think you would do well to restrict your topic much more. Either that, or present only a broad overview of the challenges presented by these fields of study.Working of artificially intelligent language translator?
lets see......Im sure I left it in the kitchen drawer.....right under my tricorder.Working of artificially intelligent language translator?
Hi,



1) Sounds are transverse pressure waves in the air. There is no noise in space, where there is no air.



2) For human speach, the waves are created by vibrations in our throat, around the vocal cords.



3) The different sounds we make are different frequencies of vibration of our vocal cords.



4) An acoustic transducer (microphone) translates pressure waves in the air (our voice) into an electric current in a ciruit.



5) The electric current is sampled, and converted into digital codes (ones and zeros).



6) A microprocessor is used to perform math functions like fourier transforms on the stream of digital codes. The Fourier transform tells the microprocessor what frequencies the transducer is detecting.



7) The processor compares patterns of frequencies to a phonetic look up table to construct words and phrases.



8) A computer program checks the word order and guesses the intended sentence structure and sorts homonyms, just like Microsoft Word or Lotus notes can correct spelling and grammar.



9) Another computer program like babelfish's translates one text language into anotyer.



10) The processor references another lookup table for pronunciations of the translated words. The pronunciations are binary coded



11) The coded sounds are sent to a DAC, or digital-to-aalog converter. The DAC translates the codes into a changing electric current.



12) A vibrating magnet in a coil of wire (speaker) translates the electric current to vibrations in the air.



13) Your ears hear the translation.