Home People Research Publications Demos
News Jobs Prospective
About Internal

Jose Costa Pereira

Department of Electrical and Computer Engineering,
University of California, San Diego
9500 Gilman Drive, Mail code 0409
La Jolla, CA 92093-0409

EBU 1, Room 4604

Email: josecp at u c s d . e d u

[Scholar profile]     [CV]

I'm currently a graduate student at the Statistical Visual Computing Lab, Department of Electrical and Computer Engineering at University of California, San Diego (UCSD). I've received a Licenciatura (5 years degree) in Computation and Systems Engineering from Faculdade de Engenharia da Universidade do Porto in 2000, and a M.A. in Computational Methods in Science and Engineering from a joint-venture of Faculdade de Ciencias and Faculdade de Engenharia da Universidade do Porto in 2003.
I've worked for Vodafone Portugal in the Data Networks support group from October 2000 until August 2005, after which I've joined the IP Division in Alcatel-Lucent from September 2005 until August 2008.


My current research is focused on image retrieval and annotation. I'm exploring contextual relations for image annotation. Context can mean several things; I've been using auxiliary sources of information for context extraction that can be used on image regularization. We argue that the use of text in image regularization is the natural thing to do since in so many interesting applications (e.g. image retrieval) text segments co-exist with images. The effectiveness of this regularization can be shown in multiple datasets to produce more accurate image retrieval systems.
In the past I've developed an Independent Component Analysis (ICA) based method for Blind Source Separation (BSS) as part of my MSc. thesis.


Graduate Fellowship, Fundacao para a Ciencia e Tecnologia, 2008 - 2012

PhD thesis:
Adaptation of Visual Models with Cross-modal Regularization
J. Costa Pereira
University of California San Diego, October 2015 [pdf] [BibTeX]

Cross-modal Domain Adaptation for Text-based Regularization of Image Semantics in Image Retrieval Systems
J. Costa Pereira, N. Vasconcelos
Computer Vision and Image Understanding
Vol. 124, pp. 123-135, July 2014 © Elsevier [ps] [pdf] [BibTeX]

On the Role of Correlation and Abstraction in Cross-Modal Multimedia Retrieval
J. Costa Pereira, E. Coviello, G. Doyle, N. Rasiwasia, G. Lanckriet, R.Levy, N. Vasconcelos
IEEE Transactions on Pattern Analysis and Machine Intelligence
Vol. 36(3), pp. 521-535, March 2014 © IEEE [ps] [pdf] [BibTeX]

Large Margin Discriminant Dimensionality Reduction in Prediction Space
M. Saberian, J. Costa Pereira, C. Xu, N. Vasconcelos
In proceedings of Advances in Neural Information Processing Systems (NIPS),
Barcelona, Spain - Dec. 2016 © NIPS [pdf] [BibTeX]

Sentiment Retrieval on Web Reviews using Spontaneous Natural Speech
J. Costa Pereira, J. Luque, X. Anguera
In proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP),
Florence, Italy - May 2014 © IEEE [ps] [pdf] [BibTeX]

On the Regularization of Image Semantics by Modal Expansion
J. Costa Pereira, N. Vasconcelos
In proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
Providence, Rhode Island, USA - Jun. 2012 © IEEE [ps] [pdf] [BibTeX]
[code] [online demo]

Maximum Covariance Unfolding: Manifold Learning for Bimodal Data
V. Mahadevan, C. Wah Wong, J. Costa Pereira, T. T. Liu, N. Vasconcelos, L. K.Saul
In proceedings of Advances in Neural Information Processing Systems (NIPS),
Granada, Spain - Dec. 2011 © NIPS [ps] [pdf] [BibTeX]

A New Approach to Cross-Modal Multimedia Retrieval
(Best student paper award)
N. Rasiwasia, J. Costa Pereira, E. Coviello, G. Doyle, G.R.G. Lanckriet, R.Levy, N. Vasconcelos
In proceedings of ACM International Conference on Multimedia (ACM-MM),
Florence, Italy - Oct. 2010 © ACM [ps] [pdf] [BibTeX]

Text and Image Modelling
  Cross-Modal Multimedia Retrieval
The problem of joint modeling text and image components of multimedia documents is studied. Two hypotheses are investigated: that 1) there is a benefit to explicitly modeling correlations between the two components, and 2) this modeling is more effective in feature spaces with higher levels of abstraction.
  Regularization on CBIR
Representation of image in semantic spaces is at the very core of many problems in Computer Vision. We provide an approach to transfer knowledge from texts associated with images in order to improve the accuracy of image semantic representations. Results are shown in the task of content-based image retrieval in three different datasets.

Last update: July, 2015