VIRTUAL PERSONAL ASSISTANT




INTRODUCTION: A computer-based virtual assistant the behavior of which can be changed by the user, comprising a voice user interface for inputting information into and receiving information from the virtual assistant by speech, a communications network, a virtual assistant application running on a remote computer, the remote computer being electronically coupled to the user interface via the communications network, wherein the behavior of the virtual assistant changes responsive to user input. 

Virtual personal assistants are artificial intelligence systems that perform tasks on a computing device in response to natural-language requests from a user. For example, a virtual personal assistant may handle opening Youtube, Google, Gmail, News, Weather, time ,etc. VPA is used to save time respect to busy life. it will easier to use and it will help a lot.

OBJECTIVE: To recognize speech and behave as per the instruction is the main objective of this assistant. As implemented Virtual personal assistant it does like as follows, after opening , it gives you greetings like "Hello" , " Good morning/afternoon/Evening" , as per the timing. it is also work like if user asked to open browser it will open the browser, then open youtube, open wikipedia , open camera ,etc. This assistant will help you in future aspect to more about your work will be going to easier.  

IMPLEMENTATION:

1. Voice recognition:

Key foundation for virtual personal assistant apps is voice recognition accuracy of user commands.

  1. Raw Speech of user must be converted from physical sound to electrical signal with device microphone, to retrieve digital data using analog-digital converter.
  2. Tokenization: In above process, the input text sequence is split into individual word called token.
  3. Identification of command to be executed is done by using conditional logic following symbolic AI.
2. Testing and Evaluation:

After adding required machine learning python libraries, range of commands are issued to get tested by developer to virtual assistant, further accuracy is noted.
The noise interference when operating on microphone should not be compromised and neglected, continuous testing is performed for further evaluation of VPA.

SNAPSHOTS: 
1. For Desktop application


Import Library

Take user command 

fig : wolframalpha is a computational Search engine API



Fig 1: Greeting


Fig 2: After saying open Youtube it will open a youtube


Fig 3: It will open google


Fig 4: After saying bye it will shutting down


Fig 5: Perform Geographical and computational task


2. For Mobile application




So here is the some snapshots of tasks which will perform by assistant. so this is how it will work.


FUTURE ENHANCEMENT: Complex machine learning techniques like natural language processing(NLP) will get added, making application more resourceful and able to handle augmented human intellect by increasing capability of VPA to solve complex problem situation, by deriving solution to problems. Implementing complex machine learning algorithms that’ll harness capability to learn past experience. For android application we can mane it more efficient. Make GUI more attractive more command can be added it can work more effectively.

REFERENCES:
  1. Imrie P, and Bednar P. (2013). ’Virtual Personal Assistant’, in: Martinez M. and Pennarolaecilia F. (editors). ItAIS 2013. Proceedings of 10th Conference of the Italian Chapter of AIS, ’Empowering society through digital innovations’, Università Commerciale Luigi Bocconi in Milan, Italy, December 14th, 2013. ISBN: 978-88-6685-007-6
  2. http://www.engpaper.com/artificial-intelligence-research-papers-2012.html
  3. https://ieeexplore.ieee.org/abstract/document/8301638/

Here, the Github Link and also a video for reference.

-   Github Link                                        
-   Video


Comments

Post a Comment