/Digital-ease

Interact with the site by using your voice!

Primary LanguageC#

Digital-ease

Capstone for Epicodus C#/.NET.
This is an idea for a hands free web-developement app. Using speech-to-text technology, the user will be able to interact with the app and build/style a simple website.

Specs/Wishlist

Sections of developement Implemented?
API call using curl or postman Postman - X
api call using RestSharp NO WAY
use the Watson Nuget Packages to successfully transcribe speech to text X
find a way for a user to record then save audio directly from the page X
output text on page X
build corpus for code-specific words/sentences O
build keyword file O
build keyword references O
build code snippets that link to a keyword O
parse returned json for transcript X
split json parse into array X
parse array for keywords X
change view based on code-snippets X
loading gif/img while user waits X
TO DO Example (if needed)
Account for different ways of structuring sentences "add color background yellow" vs "add background yellow color" etc
Add ability to create a form "create form with parameters two text fields, three number fields, one date field"
Restructure method calls to make the data path easier to follow split into if statements regarding change/add
Add buttons to allow user to revisit a successful change Change background color to green (as a button)
Persist correct statements for the user to retrieve and be able to change multiple aspects of the site at a time change bg color AND font color

Project Build

Download from the repo

Open Viual Studio 2015 and choose the Open Project option.

Navigate to the directory in which you downloaded the repo and open the project file.

Packages

Nuget Packages to ensure are installed:

  • IBM.WatsonDeveloperCloud.SpeechToText.v1 - documentation
  • Newtonsoft.Json

Watch for Reference Restoration errors, you may need to add System.Runtime.4.1.0 to the project.

Running the Server

After installation , press f5 or the play button in the toolbar to run the server and have the site appear in your default browser.

Known Bugs

Speech to text is buggy. Its almost impossible for a human to fully understand what another human says every time, so lets not assume the computer knows either. Need to account for much more variation, which will help.

Technologies Utilized

Credits

  • David Wilson - main project and idea
  • see comments in Javascript files