Modularna simulacija za testiranje propusnosti paralelnih podatkovnih tokova u sustavima za detekciju mentalnih stanja u stvarnom vremenu

master's thesis

• Title (english): Modular simulation for testing the throughput of parallel data streams in real-time mental state detection systems
• Author: Dominik Brdar
• Mentor: Siniša Popović (mentor)
• Committee member: Siniša Popović (predsjednik povjerenstva)
• Committee member: Vlado Sruk (član povjerenstva)
• Committee member: Nikolina Frid (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-12, Croatia

Abstract

Sustavi za detekciju mentalnih stanja u stvarnom vremenu imaju potencijal pružanja potpore u visokostresnim zanimanjima gdje je mentalno stanje osoba od kritične važnosti. Primjena takvih sustava u ograničenim radnim okruženjima, izvođenje u stvarnom vremenu i složenost neurobiologije predstavljaju razne izazove. U ovom radu ću predstaviti pregled literature u području računalne detekcije mentalnih stanja na osnovu psihofizioloških signala, a zatim i karakterističnu arhitekturu takvih sustava, s naglaskom na izvođenje u stvarnom vremenu. Stoga ću na osnovu predstavljene arhitekture oblikovati modularnu simulaciju takvih sustava, a zatim i programski ostvariti dizajn za integraciju modula i testiranje potrebnih propusnosti paralelnih podatkovnih tokova na heterogenom računalnom sustavu. Konačni rezultati pokazuju da ostvarena simulacija izvođena na stolnom računalu opremljenom grafičkom karticom pruža potrebnu paralelizaciju i dostatne propusnosti podatkovnih tokova u svrhu zadovoljavanja ograničenja rada u stvarnom vremenu koja je realno očekivati u ovoj domeni primjene.

Abstract (english)

Real-time mental state detection systems have the potential to provide support in high-stress professions where individuals' mental states are critically important. The application of such systems in constrained work environments, real-time execution, and the complexity of neurobiology present various challenges. In this paper, I will present a literature review in the field of computer-based mental state detection using psychophysiological signals, followed by a description of the typical architecture of such systems, with emphasis on real-time execution. Therefore, based on the presented architecture, I will design a modular simulation of such systems and then implement a software design to integrate the modules and test the required throughput of parallel data streams on a heterogeneous computer system. The final results demonstrate that the implemented simulation, performed on a desktop computer equipped with a graphics card, provides the necessary parallelization and sufficient data stream throughput to satisfy the real-time constraints that may be expected in this application domain.

• Keywords: simulacija; detekcija mentalnih stanja; stvarno vrijeme; propusnost podatkovnih tokova; heterogeni računalni sustav; grafička kartica; procesor za digitalnu obradu signala; OpenCL
• Keywords (english): simulation; mental state detection; real time; data stream throughput; heterogeneous computing system; graphics card; digital signal processor; OpenCL
• Language: croatian
• URN:NBN: urn:nbn:hr:168:018055
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 08:32:26