Desktop-AI-Assistant
This is an AI assistant which takes speech as input and performs various tasks
Working
- The code is run and the AI assistant greets the user with a welcome message.
- It displays the input audio as text in the terminal and performs the directed task.
- If the assistant fails to understand the given command, then it displays the text on the terminal.
STEPS TO RUN THIS PROJECT
- Download the main.py file and place it in the project folder of the desired name.
- Install the required packages.
- pyttsx3
- SpeechRecognition
- PyAudio
- openai
- pywin32
- wheel
- Run the code.
- Speak 'close' to exit.
Tasks that the AI can perform
- Searching content in Wikipedia - the user needs to include the word - 'Wikipedia' to get the desired result. ** For example - what is a computer Wikipedia?
- Launching youtube and Google websites -the user needs to include the word - 'open youtube' or 'open Google' in his speech to get the desired result.
- Displaying images in a specified folder -the user needs to include the word- 'show images' in his speech to get the desired result.
- Displaying the current time - the user needs to include the word- 'the time' in his speech to get the desired result.
- Launching VS Code editor -the user needs to include the word - 'open code' in his speech to get the desired result.
- Mailing -the user needs to include the word - 'email' followed by telling the content and receiver's email address in his speech to get the desired result.
Libraries used
- pyttsx3 library for text-to-speech conversion
- speech_recognition library for performing speech recognition
- datetime library for date and time
- wikipedia library to access and parse data from Wikipedia
- webbrowser library to access web browser
- os library for using operating system dependent functionality
- smtplib library to send mail
API used
- Google Web Speech API - for recognising voice input
- Microsoft Speech API (SAPI5) - for voice recognition and synthesis