Technology: Arduino, Heroku, ESP Microcontroller

Description: This system was developed to monitor the sound intensity levels of an environment so that notifications could be issued if the sound levels pass a user-specified threshold. The system is contained within an IP-67 certified enclosure ensuring indoor and outdoor usage.

Technical Details: A web API was created to interface the microcontroller to the PostgrSQL database. This allowed the intensity levels to be tracked over a period of time to eventually be visualized via a front-end application. A power distribution board was used to convert a typically available 110V AC source to 3.3V and 12V DC. An ESP-01s was interfaced with an industrial grade noise decibel detection module.

Technology: TensorFlow, Keras, PyPI, HuggingFace

Description: The emotion detection model is part of a journal therapy mobile application that will be developed soon call IRIS. The model is used to detect emotional sentiments in sentences to determine how depressed an individual might be. An API for the model was developed and launched through Heroku and a Python Package was created called iris-emotion through PyPI

Technical Details: An RNN model was used for the emotion detection. The model is capable of accurately recognizing up to 6 emotions. The model was trained using a combination of the ISEAR dataset and a HuggingFace emotions dataset.