Očitavanje pulsa osobe iz video snimke

undergraduate thesis

• Title (english): Estimating the Person's Pulse From a Video Recording
• Author: Ivan Juraga
• Mentor: Marko Subašić (mentor)
• Committee member: Marko Subašić (predsjednik povjerenstva)
• Committee member: Ivana Zadro Matovinović (član povjerenstva)
• Committee member: Karlo Koščević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Završni rad je bila primjena dubokih neuronskih mreža u svrhu određivanja pulsa osobe preko video snimke lica. Izrada rada sastojala se od 3 glavna dijela: sakupljanje video zapisa i izrada skupa podataka, programiranje samog modela te treniranje modela u svrhu veće točnosti. Modelova mreža je sastavljena od kombinacije konvolucijske, ponavljajuće i potpuno povezane mreže sa adam optimizatorom i MAE kao načinom izračuna pogreške. Model trenutno može sa sigurnošću pretpostaviti da li se radi o niskom, srednjem ili visokom pulsu, a loss mu trenutno iznosi =~4.9 što je uzeći u obzir nepreciznost instumenta za izračun pulsa zadovoljavajuće . Sama implementacija je izvedena u programskom jeziku Python koristeći biblioteke Tensorflow i Keras.

Abstract (english)

Thesis based predicting pulse by monitoring person's face using deep neural network. The making of the thesis consists of three large parts: making a dataset, programing and finally training the model. The dataset was made specially for this thesis and is still not finished, while the model is made as a combination of convolutional,fully connected and reccurent network with mean square error as the loss function and adam as the optimizer. Right now the model gives decent results but the model will be further worked on to try prove that you can predict pulse from a video of a person's face. The code is written in python with Tensorflow and Keras as the driving force of the model.

• Keywords: duboka neuronska mreža; puls osobe; video snimka lica; Python; Tensorflow; Keras
• Keywords (english): Deep Learning; Neural Network; Pulse; Face Pulse; Tensorflow; Keras
• Language: croatian
• URN:NBN: urn:nbn:hr:168:811076
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 18:47:57