Virtual Assistant

Question

Good afternoon fellows, as you go?

So I’m fairly new to this development environment and I programmed in python some time ago. I have a project idea where it involves creating a kind of virtual assistant (a child version of a Siri or Amazon Alexa). I’ve done some research, seen some videos and put together some libraries and frameworks that I believe will help me with this project. However, I would like to know whether these libraries will not conflict in any way and whether this is the best way I can go.

Another question I have is about python UI. The best choice today is Kivy? And it will be possible that I can develop this code, with these libraries with python kivy to create an Android app?

The libraries/frameworks are these:

Voice recognition - https://pypi.org/project/SpeechRecognition/

Voice synthesizer - https://pypi.org/project/pyttsx3/

Chat Bot - https://chatterbot.readthedocs.io/en/stable/

Search (Mining) on the Internet - https://scrapy.org

Graphical interface + Cross-Platform (Android?) - https://kivy.org/#home // http://excript.com/python/kivy.html

I accept tips, opinions and insults. I appreciate the understanding of all!

Answer 1

In Python we have an amazing framework to build native applications that can be packaged for these various platforms! So if you want to develop for Android, iOS, Windows, Linux or Mac OS using our beloved Python, Kivy is a great bet! Other advantages of Kivy: it is totally free, is already in its stable versions and is very well documented. All these libraries you’ve listed are very good. These links might help you https://www.treinaweb.com.br/blog/criando-um-chatbot-com-python/

and take a look at these other libraries: Speech Recognition Library, Speech Synthesis with Pyttsx3

Answer 2

I’m working on something similar right now, however I couldn’t streamline my application using speech_recognition from python, as an alternative found some other libraries/programs that can help you if you are not satisfied with the performance of speech_recog, they are Snowboy and picovoice do similar tasks but with the advantage of being offline and much faster, the only addendum I make to you about them is that they have very little material and are (at least for me) very difficult to install were days of research to solve all the mistakes that happen until they start to work.

How you want to use on Android picovoice already comes with a demo "ready" for you to test.

Here is their link:

https://snowboy.kitt.ai/

https://picovoice.ai/