LogicAnalyzer

Pi Audio Device

pi + rainbowhat + touchscreen + usb sound card (mic or aux in) + open ai = audio logic analyzer

LogicAnalyzer TwitterBot

LogicAnalyzer DiscordBot

still under development

NEW web version

web versions upload data to nodejs app running on vps

button operated interview mode and auto mode with aux in that will detect 2 second pause to end recording and analyze. Audio samples should be with less than roughly 3 minutes, 4 if its a slow speeker.

Main Page

Web Analysis (all files)

Screen version

It will listen for speech, when speech stops (or paused) it will analyze what was said with whisper+gpt3.5turbo, it will display # of LF on the display, leds represent how much of what was said was fallacious and it will print its explanations on the touch screen OR hdmi as tty1.

If no fallacies are found

If fallacy found

BS METER INSTALL INSTRUCTIONS

sudo apt update

raspi-config enable i2c

sudo nano /usr/share/alsa/alsa.conf

Look for these lines:

defaults.ctl.card 0

defaults.pcm.card 0

Replace with:

defaults.ctl.card 1

defaults.pcm.card 1

sudo apt install git python3-pip python3-pyaudio python3-smbus

git clone https://github.com/matt-desmarais/LogicAnalyzer.git

pip3 install pyaudio rainbowhat openai SpeechRecognition pydub