Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

You searched for id:"oai:www.bdigital.unal.edu.co:64612". One record found.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters


Universidad Nacional de Colombia

1. Vanegas Ramírez, Jorge Andrés. Large-scale non-linear multimodal semantic embedding.

Degree: http://bdigital.unal.edu.co/64612/, 2018, Universidad Nacional de Colombia

The main goal of this thesis is to investigate effective and efficient methods to combine complementary evidence, and model the relationships between multiple modalities of multimedia data in order to improve the access and analysis of the information, to finally obtain valuable insights about the data. In this thesis is proposed to use multimodal latent semantic as the strategy that allows us to combine and to exploit the different views from this heterogeneous source of knowledge, by modeling relations between the different modalities and finding a new common low-dimensional semantic representation space. For a richer modeling, it is proposed the usage of kernel-based methods that usually present accurate and robust results. Unfortunately, kernel-based methods present a high computational complexity that makes them infeasible for large data collections. This drawback implies one of the most important challenges addressed in this thesis, which was to investigate alternatives to handle large-scale datasets with modest computational architectures. In this thesis, several kernelized semantic embedding methods based on matrix factorization have been proposed, developed and evaluated. Thanks to the non-linear capabilities of the kernel representations, the proposed methods can model the complex relationships between the different modalities, allowing to construct a richer multimodal representation even when one of the modalities presents incomplete data. Besides, the proposed methods have been designed under a scalable architecture based on two main strategies: online learning and learning-in-a-budget that allow preserving low computational requirements in terms of memory usage and processing time. An extended experimental evaluation shows that the proposed multimodal strategies achieve the state-of-the-art in several data analysis tasks, such as multi-labeling and multi-class classification and cross-modal retrieval and under different learning setups, such as supervised, semi-supervised, and transductive learning. Furthermore, thanks to the online learning and learning-in-a-budget strategies proposed in this thesis, the scalability capabilities are preserved allowing to deal with large-scale multimodal collections.

Resumen: El objetivo principal de esta tesis es investigar m´etodos eficaces y eficientes para combinar evidencia complementaria de múltiples modalidades de información multimedia y modelar las relaciones entre éstas, con el fin de mejorar el acceso y el análisis de los datos contenidos. En esta tesis se pretende utilizar la estrategia de semántica latente multimodal, la cual permite combinar y explotar las diferentes vistas de esta fuente de información heterogénea, modelando las relaciones entre las diferentes modalidades y encontrando un nuevo espacio com´un de representación semántica de baja dimensionalidad. Para un modelado más rico, se propone el uso de métodos basados en kernel los cuales usualmente presentan resultados precisos y robustos. Desafortunadamente, los métodos basados en…

Subjects/Keywords: 0 Generalidades / Computer science, information & general works; 62 Ingeniería y operaciones afines / Engineering

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Vanegas Ramírez, J. A. (2018). Large-scale non-linear multimodal semantic embedding. (Thesis). Universidad Nacional de Colombia. Retrieved from http://bdigital.unal.edu.co/64612/1/doctoral-thesis-jorge.pdf

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Vanegas Ramírez, Jorge Andrés. “Large-scale non-linear multimodal semantic embedding.” 2018. Thesis, Universidad Nacional de Colombia. Accessed December 13, 2018. http://bdigital.unal.edu.co/64612/1/doctoral-thesis-jorge.pdf.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Vanegas Ramírez, Jorge Andrés. “Large-scale non-linear multimodal semantic embedding.” 2018. Web. 13 Dec 2018.

Vancouver:

Vanegas Ramírez JA. Large-scale non-linear multimodal semantic embedding. [Internet] [Thesis]. Universidad Nacional de Colombia; 2018. [cited 2018 Dec 13]. Available from: http://bdigital.unal.edu.co/64612/1/doctoral-thesis-jorge.pdf.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Vanegas Ramírez JA. Large-scale non-linear multimodal semantic embedding. [Thesis]. Universidad Nacional de Colombia; 2018. Available from: http://bdigital.unal.edu.co/64612/1/doctoral-thesis-jorge.pdf

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

.