Adapted Vocabularies for Generic Visual Categorization
- F. Perronnin, C. Dance, G. Csurka, M. Bressan
- Computer ScienceEuropean Conference on Computer Vision
- 7 May 2006
A novel and practical approach to GVC is proposed, which describes the content of all the considered classes of images, and class vocabularies obtained through the adaptation of the universal vocabulary using class-specific data.
Nonparametric discriminant analysis and nearest neighbor classification
- M. Bressan, Jordi Vitrià
- Computer SciencePattern Recognition Letters
- 1 November 2003
Tone-mapping high dynamic range images by novel histogram adjustment
- Jiang Duan, M. Bressan, C. Dance, G. Qiu
- Computer Science, Environmental SciencePattern Recognition
- 1 May 2010
Choose the Damping, Choose the Ranking?
- M. Bressan, E. Peserico
- Computer ScienceWorkshop on Algorithms and Models for the Web…
- 12 February 2009
It is shown that rank is indeed relatively stable in both graphs, and that “classic” PageRank marginally outperforms Weighted In-degree, mainly due to its ability to ferret out “niche” items.
A weighted non-negative matrix factorization for local representations
- David Guillamet, M. Bressan, Jordi Vitrià
- Computer ScienceProceedings of the IEEE Computer Society…
- 8 December 2001
An improvement of the classical Non-negative Matrix Factorization (NMF) approach for dealing with local representations of image objects by incorporating prior knowledge in the form of a weight matrix extracted from the training data.
Motivo: Fast Motif Counting via Succinct Color Coding and Adaptive Sampling
- M. Bressan, S. Leucci, A. Panconesi
- Computer ScienceProceedings of the VLDB Endowment
- 4 June 2019
The novel tools for the `motif counting via color coding' framework are developed, including an adaptive graphlet sampling strategy that breaks the additive approximation barrier of standard sampling, and a biased coloring trick that trades accuracy versus running time and memory usage.
BBVA's Data Monetization Journey
- Elena Alfaro, M. Bressan, F. Girardin, J. Murillo, I. Someh, B. Wixom
- Economics, Computer ScienceMIS Q. Executive
- 1 June 2019
Based on BBVA’s efforts, a rich data monetization portfolio is developed by making preexisting data monetizing activities more effective, by pursuing new data monetized approaches, and by investing in a larger number ofData monetization projects.
Using an ICA representation of high dimensional data for object recognition and classification
- M. Bressan, David Guillamet, Jordi Vitrià
- Computer ScienceProceedings of the IEEE Computer Society…
- 8 December 2001
A Bayesian classification scheme to the problem of recognition through probabilistic modeling of high dimensional data is applied and how this approach outperforms other techniques commonly used in the context of appearance-based recognition is shown.
Inferring the demographic history of the Adriatic Flexopecten complex.
- J. Pujolar, T. Marčeta, C. Saavedra, M. Bressan, L. Zane
- BiologyMolecular Phylogenetics and Evolution
- 1 November 2010
Bayesian Classification of Cork Stoppers Using Class-Conditional Independent Component Analysis
- Jordi Vitrià, M. Bressan, P. Radeva
- Computer ScienceIEEE Transactions on Systems Man and Cybernetics…
- 2007
A class-conditional independent component analysis representation of the data is proposed that allows an accurate estimation of theData probability density function by factorizing it.
...
...