The Pillow Project

A Brown HCI Group Research Project

Prototype developed by Tiffany Chen as her senior thesis

Dependencies

• Python 3
• Arduino
• SpeechRecognition
• PyAudio
• Portaudio
• FLAC
• Matplotlib
• scikit-learn,
• espeak
• numpy
• scipy

Hardware

• Raspberry Pi 3 Model B+
• Arduino Uno
• Adafruit BNO055
• FSR

Scripts

• prototype_voiceRecognition.py

  • Uses Google Speech Recognition API to listen to initialization commands.
  • Triggers calibration script.

• prototype_model_data_collector.py

  • Calibration Script. Collects data for model training.

• prototype_sleep_data_collector.py

  • Collects data during sleep. Needs to be started manually.

• prototype_classifier_decisiontree.py

  • Trains the model with data collected during calibration. The model is then used to classify data collected during sleep.
  • Can be changed to use other machine learning models.

• prototype_sendData.py

  • Sends data files to SleepCoacher server.

To Run

• Run prototype_voiceRecognition.py to use Voice User Interface (VUI).

• Run prototype_model_data_collector.py to calibrate the pillow (train machine learning model).

• Run prototype_model_data_collector.py to collect data during sleep.