Skip to content

A chat bot with conversation driven learning capabilities, and computer vision

Notifications You must be signed in to change notification settings

ericpelland/E.V.A.

Repository files navigation

E.V.A.

(Eric's Virtual Assistant)

Voice Recognition

  • Talk to E.V.A. and she will listen and decipher what you are saying

Voice Synthesis

  • E.V.A. responds back with a synthesized voice

Command and Trigger learning

  • E.V.A. can learn new trigger phrases, commands and responses all without touching the code.
  • Just tell her you want to add a command or trigger

Chatbot

  • E.V.A. can utilize the Dice coefficient statistic and the Levenshtein distance metric to be a bit of a chatbot. Some Easter eggs and jokes added in as well.

Trigger Phrases

  • Tell her to stop using trigger words, and she will start listening and responding without the need to use her name.
  • You can turn trigger words back on by telling her to.

Giphy

  • E.V.A. has a giphy integration and displays a personable giphy face
  • E.V.A. can respond by showing gif's
  • She can also change her appearance upon request.

Youtube

  • Watch any video on youtube just by asking
  • E.V.A. can search and embed the youtube video right into her (inter)face.
  • She can also close the video upon request.

Webcam Control

  • Eva can open her eyes by enabling your webcam. Just ask.
  • If you want to turn off her webcam, as usual, just ask.

Computer Vision

  • E.V.A. can use the MobileNet classification model with ML5 and a TensorFlow back to classify images from video
  • Ask E.V.A. what she is looking at after enabling the webcam, and she will try her best!

Dictionary

  • Get the definition of any word

Conversation and Command Responses:

  • E.V.A. has pre-programmed(but can still be dynamic) responses.
  • The path from your input to the chatbots response is AI assisted with natural language processing technologies
  • Responses can be single responses to a phrase, or a random selection of many available responses.
  • Procedural tasks can take in a series of responses and functions to complete one after another

Command Examples:

  • 'eva are you listening'
  • 'eva can you hear me'
  • 'hi eva add trigger'
  • 'hello eva add command'
  • 'hey eva show me a telescope'
  • 'eva can you dance'
  • 'hi eva change your face'
  • 'eva show me a video of a monkey'
  • 'eva exit youtube'
  • 'eva define easter'
  • 'eva what are you up to'
  • 'eva tell me a joke'
  • 'eva what does the fox say'
  • 'eva how are you today'
  • 'eva can you disable vision'
  • 'eva open your eyes'
  • 'eva what are you looking at'
  • 'eva what do you see'
  • 'eva disable trigger words'
  • 'eva stop listening'
  • 'eva show me what you can see'
  • 'eva hide webcam stream'
  • 'eva what are you looking at'

Install

  • git clone https://github.com/ericpelland/E.V.A./
  • cd E.V.A./
  • npm install
  • update youtube(google), and dictionary(oxford) API keys
  • export YOUTUBE_API_KEY="API KEY HERE"
  • export DICTIONARY_APP_ID="API KEY HERE"
  • export DICTIONARY_API_KEY="API KEY HERE"
  • node index.js

About

A chat bot with conversation driven learning capabilities, and computer vision

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published