Teammate 1: Tonderai Kambarami

Teammate 2: Luis Guillermo Usseglio-Carbajal

Teammate 3: Reed Bader

Teammate 4: Tarik Salay

---

Introduction

The objective of this lab was to build a raspberry pi based program to detect keywords from live streaming audio. The system will detect the keyword and display how many times it was repeated on the screen. Our system didnt read words from a wav file but rather live audio from a microphone. Completing the bonus

Objectives

• Build a raspberry pi based program for speech recognition
• Allow for user to prompt keywords to parse
• Allow for live audio streams from microphones to be determined

Approaches/Methods

We used python libraries to complete this LAB. This allowed us to develop the application on a laptop and port it to the raspberry pi. This is because all we needed to do was to install the libraries on python and run the code through Thonny IDE. Speech to text is very straightforward when utilizing Speech_Recognition and Pyaudio libraries so development was trivial. Porting to the raspberry pi was also trivial since the Speech_Recognition and Pyaudio libraries are native to the raspberry pi. All that is needed is a compatible microphone plugged into the raspberry pi and to run the program from the command line.

Workflow

• Develop Program on Pycharm IDE on PC
• Port program to Raspberry Pi
• Run program on Raspberry Pi

Evaluation

Below you will find a image of our code running

LAB Video

Lab 2 video

Conclusion

This lab was very straightforward, all that was required was to write a script using the python libraries which can be easily ported to the raspberry pi. They can be run natively on the raspberry pi through the IDE with an microphone plugged into the USB.