Využití velkých jazykových modelů a počítačového vidění pro sledování energetického příjmu a výdeje v mobilní aplikaci

Abstract

This bachelor thesis focuses on the design, implementation, and experimental verification of the FoodieStats mobile application, which utilizes Large Language Models (LLMs) to track energy intake and expenditure. The primary objective of the work is to reduce user friction associated with manual dietary logging through the analysis of text and image inputs. To achieve this goal, a system based on a client-server architecture was designed, comprising a native Android application developed in Kotlin and a backend API built using the Spring Boot framework.