Automatic lip synchronization by speech signal analysis

scientific master's theses

• Title (croatian): Automatska sinkronizacija usana pomoću analize govornog signala
• Author: Goranka Zorić
• Mentor: Igor Sunday Pandžić (mentor)
• Committee member: Maja Matijašević (predsjednik povjerenstva)
• Committee member: Igor Sunday Pandžić (član povjerenstva)
• Committee member: Robert Manger (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; (Department of Telecommunications); Zagreb
• Defense date and country: 2005-10-21, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering
• Universal decimal classification (UDC): 621.3 - Electrical engineering ; 004 - Computer science and technology. Computing. Data processing

Abstract

This master thesis investigates automatic lip synchronization. It is a method for generating an animation of 3D human face model where the animation is driven only by a speech signal. The whole process is completely automatic and starts from the speech signal. The automatic lip synchronization consists of two main parts: audio to visual mapping and a face synthesis. The thesis proposes and implements a system for the automatic lip synchronization of synthetic 3D avatars based only on the speech input. The speech signal is classified into viseme classes using neural networks. The topology of neural networks is automatically configured using genetic algorithms. Visual representation of phonemes, viseme, defined in MPEG-4 FA, is used for face synthesis. The system is adopted for specificity of the Croatian language. Detailed system validation based on three different evaluation methods is done and potential applications of these technologies are discussed in details. This method is suitable for real-time and offline applications. It is speaker independent and multilingual.

Abstract (croatian)

Ovaj magistarski rad istražuje automatsku sinkronizaciju usana. To je metoda generiranja odgovarajuće animacije 3D modela ljudskog lica s obzirom na postojeći zvučni signal govora, tako da se dobiva animirani lik koji govori. Postupak je potpuno automatiziran i počinje od zvučnog signala govora. Automatska sinkronizacija govora sastoji se od dva dijela: preslikavanja iz ulaznog audio u izlazni vizualni skup parametara i sinteze lica. U radu je predložen i napravljen sistem za automatsku sinkronizaciju usana temeljen samo na govornom signalu. Klasifikacija govornog signala u grupe vizeme izvedena je pomoću neuronskih mreža. Optimalna topologija neuronskih mreža se automatski određuje pomoću genetičkih algoritama. Vizemi, vizualni predstavnici fonema, definirani su u MPEG-4 FA standardu, a koriste se za sintezu lica. Sistem je optimiziran za specifičnosti hrvatskog jezika. Za validaciju sistema korištene su tri različite metode testiranja., a potencijalne primjene ovakvih tehnologija izložene su do detalja. Opisana metoda za sinkronizaciju usana prikladna je za primjene u stvarnom vremenu ali i za offline produkcije. Neovisna je o govorniku i višejezična je.

• Keywords: Lip synchronization; facial animation; MPEG-4 FBA; virtual characters; speech processing; neural networks; genetic algorithms
• Keywords (croatian): Sinkronizacija usana; animacija lica; MPEG-4 FBA; virtualni ljudi; obrada govora; neuralne mreže; genetički algoritmi
• Language: english
• URN:NBN: urn:nbn:hr:168:946018
• Study programme: Title: Postgraduate master programme in electrical engineering; Study programme type: university; Study level: postgraduate; Academic / professional title: Magistar znanosti elektrotehnike (Magistar znanosti elektrotehnike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-07-04 07:21:57