Class of 2026Salt Lake City, Utah
AboutI’d like to explore the following areas under the guidance of my mentor in the summer of 2023. - Real time path and hazard detection (Using a GPS sensor with image assistance (?) "image to location" process) - Real time path finding + navigation tech (constantly calculated most efficient path - in and out of resort. Probability detection to find average route difficulty.) - Mobile/micro-controller based front end offline AI - Holographic display basics - Ski patrol panic button My goals are 1. Complete an award-worthy high school science fair project; 2. Write an intermediate level research paper because this would be my first research page and 3. If applicable, present at a conference. I am a member of the school's speech and debate team and public speaking is my strong suit.
- "Silver Sync - The Application of Generative Pre-Trained Transformer Natural Language Processing (NLP-GPT) Based Intelligent Assistance Technology (IAT) on Seniors to Improve Their Safety, Increase Senior Autonomy and Decrease Social Isolation" with mentor Clark (Dec. 15, 2023)
Silver Sync - The Application of Generative Pre-Trained Transformer Natural Language Processing (NLP-GPT) Based Intelligent Assistance Technology (IAT) on Seniors to Improve Their Safety, Increase Senior Autonomy and Decrease Social Isolation
Started May 3, 2023
Abstract or project description
Seniors living alone or in a nursing home are often isolated from interpersonal interaction - be it chatting, assistance or responding to extreme situations. This leaves them in a vulnerable situation which continues to deteriorate with the increase of the aging population and stagnation of available support resources. To combat this, I aim to create a system based on cutting-edge Natural Language Processing (NLP) Generative Pre-Trained Transformer (GPT) to naturally and appropriately respond to the needs of a senior. To begin, an Automatic Speech Recognition (ASR) system parses speech from the user to determine their sentiment. Simultaneously, a Face-mesh Computer Vision (CV) system captures images of the user to identify their position, sentiment and potential needs (like the wish to interrupt). This data is sent to a cloud storage-base and automatically processed by a State Machine to identify the state of the senior and react accordingly. The state of the user as well as what they said is piped to a GPT that returns a text response and keeps a transcript for future conversational reference. This response is then output as speech via a Speech-To-Text (STT) system to the senior in a conversational way which allows them to respond.