Inspiration
In the present world different people speak different languages. Learning and understanding each language is difficult as well as time consuming. So, in-order to solve this problem we are having language translator apps. Most translators that we find on web are used to translate one language to another, then what if a person speaks a bilingual conversation…? It translates one language and leaves the unrecognized words unchanged. This problem occurs while translating a real-time conversation as we use words of different languages when we speak.
What it does
This project helps to translate a multi-lingual sentence into a single language which helps a lot for tourists, foreigners.
How we built it
Below I have explained the significant steps of the algorithm.
INPUT: The user needs to give input through microphone in the form of audio or the user can give the input in the form of text.
PROCESS:
- If the input type is audio, convert the given audio into text.
- Tokenize the given input and analyze number of English words and non-English words.
- Let us consider Telugu as our intermediary language (if English words are more) and English as our intermediary language (if non-English words are more).
- If the sentence has more English words, first translate the entire sentence into intermediary language followed by desired language.
- If the sentence has more non-English words, first translate the entire sentence into intermediary language followed by desired language.
OUTPUT: Output the translated sentence in the form of audio.
Challenges we ran into
As the time was limited we could only add few languages but we promise that in future we will definitely add many other features and make our translator support all national and international languages.
Accomplishments that we're proud of
All of us learnt many new interesting concepts. We were able to code the entire project on our own which we are very proud of.
What we learned
We have gained vast experience by doing this project which will help us in our future carrer path.
What's next for CODE_MIXING_TRANSLATOR
This translator supports only 3 languages(i.e Telugu,Hindi,English), in future we will add more languages which would help all kinds of people to translate different kinds of sentences. We will also improve the accuracy of our model.
Log in or sign up for Devpost to join the conversation.