Speech to Text Applications and how you can make one

Speech to text is one of the best inventions of modern day technology. There are a number of times we want to take notes and doing just a voice note by recording the sound doesn't make it convenient to identify the content with just a look. 
Text to speech to the rescue. Now, you can not only record your voice but actually just convert your voice to text and be saved to your device or cloud storage for easier access. 

There are already many applications out there that do the said job and of course, the best ones are made by notable companies like Google, Microsoft, IBM, Amazon etc.

But then what about all those other Speech to Text applications out there? How are they different?
To put it in simpler words, they're not.
You could get a somewhat better result in one of the application compared to the other and that'll be only because it suits your voice better. And partially because some applications might be using machine learning to improve their word prediction.
But then so are Google and other major companies doing. You cannot possibly get a better result because the other applications are also using the same API in their back-end.

Google, Amazon, IBM and many others provide Text to speech for a very nominal fee.
In case you're developing a text to speech program you might like to look into Watson Speech to Text by IBM, Amazon Transcribe, Cloud Speech-to-Text by Google and Microsoft's Azure Speech Transcription.
Now all these Speech to Text are great and charge a minor fee. If you're looking a for a way to quickly create one Speech to Text application there are many great assets in the Unity Asset Store to give you head start.

No comments :

Post a Comment