|
Computer Graphics and Video Features for Speaker Recognition
Fér, Radek ; Matějka, Pavel (referee) ; Černocký, Jan (advisor)
We describe a non-traditional method for speaker recognition that uses features and algorithms used mainly for computer vision. Important theoretical knowledge of computer recognition is summarized first. The Boosted Binary Features are described and explored as an already proposed method, that has roots in computer vision. This method is evaluated on standard speaker recognition databases TIMIT and NIST SRE 2010. Experimental results are given and compared to standard methods. Possible directions for future work are proposed at the end.
|
|
Text Dependent Speaker Verification
Fux, Jan ; Glembek, Ondřej (referee) ; Matějka, Pavel (advisor)
The goal of this Bachelor's thesis was to design text dependent speaker recognition system. There were few systems tested for MIT database. This database contains recordings of 0.46s average length. Best case for recognition is to use a combination of DTW system using posterior probability estimation (posteriograms) as an output of Phoneme recognizer and acoustic SID system based on iVectors and PLDA (Probabilistic Linear Component Analysis). Fusion with Neural network gives the best results (EER). These are 17.84% EER for women and 16.38% for men. It's 49.9% relative improvement for women and 54.2% for men against acoustic recognition alone.
|
|
Automatic Creation of Dictionaries from Translations
Svoboda, František ; Matějka, Pavel (referee) ; Smrž, Pavel (advisor)
Goal of this thesis is to implement system, capable of extracting bilingual dictionaries from parallel texts. Reader may find examples of how to obtain such documents and description of steps leading to successfull acquirement of desired information. Mainly statistical machine translation methods were examined and used for this purpose. Besides description of created system, short analysis of problems linked with the subject can be found as well as evaluation of results.
|
|
Web as a Source for Automatic Creation of Morphological Dictionary
Bulka, Pavol ; Matějka, Pavel (referee) ; Smrž, Pavel (advisor)
Creation of natural language words is based on rules, which are generally complex. Often it is very difficult or even impossible to describe them precisely in a formal way. That is why we use a morphological dictionary to process natural language. In this paper we discuss the creation of morphological dictionary from Slovak's top level domain web. We talk about web crawling, data processing for morphological analysis and data structures too. This document makes basic principle and conception of morphological analysis clear. Final system, which is described in this thesis, produces morphological dictionary. This dictionary can be use in various application, for example spell checker, machine translation and so on.
|
| |
|
Home Information System
Knyrevich, Michail ; Matějka, Pavel (referee) ; Szőke, Igor (advisor)
This bachelor's thesis involves the development of the Informational system for home, especially its Energy module. Some existing informational systems of this topic are analyzed there. Further, there is a formal concept of self Energy module , which is based on gained information. The part of this concept is prediction system of energetic expenses, which uses a weather monitoring. Realization and testing with usage of web technologies (especially the Zend Framework) are placed at the end of my thesis.
|
|
Speaker Diarization
Tomášek, Pavel ; Karafiát, Martin (referee) ; Matějka, Pavel (advisor)
This work aims at a task of speaker diarization. The goal is to implement a system which is able to decide "who spoke when". Particular components of implementation are described. The main parts are feature extraction, voice activity detection, speaker segmentation and clustering and finally also postprocessing. This work also contains results of implemented system on test data including a description of evaluation. The test data comes from the NIST RT Evaluation 2005 - 2007 and the lowest error rate for this dataset is 18.52% DER. Results are compared with diarization system implemented by Marijn Huijbregts from The Netherlands, who worked on the same data in 2009 and reached 12.91% DER.
|
|
Automatic Data Recording of Digital Satellite Broadcast
Řezníček, Ivo ; Matějka, Pavel (referee) ; Szőke, Igor (advisor)
This work aims at the creation of a~system of massive recording of multimedia data, especially speech data in various languages. The first issue is to find out high quality data source, the second is to build the system for managing and storing received data in the digital form. A digital satellite transmission is chosen as a signal source (DVB-S system). Main system features include recording of multiple streams in parallel, support of multiple cards, retrieving and storing of additional information (from Internet) and scheduling of recordings. The system will provide massive amounts of data for training of a language identification system.
|
|
Speaker Recognition Based on Long Temporal Context
Fér, Radek ; Matějka, Pavel (referee) ; Černocký, Jan (advisor)
Tato práce se zabývá extrakcí vhodných příznaků pro rozpoznávání řečníka z delších časových úseků. Po představení současných technik pro extrakci takových příznaků navrhujeme a popisujeme novou metodu pracující v časovém rozsahu fonémů a využívající známou techniku i-vektorů. Velké úsilí bylo vynaloženo na nalezení vhodné reprezentace temporálních příznaků, díky kterým by mohly být systémy pro rozpoznávání řečníka robustnější, zejména modelování prosodie. Náš přístup nemodeluje explicitně žádné specifické temporální parametry řeči, namísto toho používá kookurenci řečových rámců jako zdroj temporálních příznaků. Tuto techniku testujeme a analyzujeme na řečové databázi NIST SRE 2008. Z výsledků bohužel vyplývá, že pro rozpoznávání řečníka tato technika nepřináší očekávané zlepšení. Tento fakt diskutujeme a analyzujeme ke konci práce.
|
|
Modelling Prosodic Dynamics for Speaker Recognition
Jančík, Zdeněk ; Fapšo, Michal (referee) ; Matějka, Pavel (advisor)
Most current automatic speaker recognition system extract speaker-depend features by looking at short-term spectral information. This approach ignores long-term information. I explored approach that use the fundamental frequency and energy trajectories for each speaker. This approach models prosody dynamics on single fonemes or syllables. It is known from literature that prosodic systems do not work as well the acoustic one but it improve the system when fusing. I verified this assumption by fusing my results with state of the art acoustic system from BUT. Data from standard evaluation campaigns organized by National Institute of Standarts and Technology are used for all experiments.
|