Language Recognition Using Multimodal Deep Learning

Student: Dymchenko Sofya

Supervisor: Ekaterina Artemova

Educational Programme: Mathematics (Bachelor)

Year of Graduation: 2019

In this work, a problem of audio and video data classification for rare languages of Russia is researched. To develop machine learning models we train and test convolutional and recurrent neural networks on specially collected dataset. First, we build a pipeline of collecting new data from local TV-shows from different regions of Russia. Second, we build deep learning models based on RNNs and CNNs to analyze and classify the collected data into several languages. We provide experiments for audio and video modalities and show that proposed methods allow to improve random predictions significantly which means that language features from audio and video can be learned by neural networks.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses