Code
- Incremental Image Alignment Library in C++. In this library we have implemented in C++, using OpenCV, our papers: [Buenaposada02b] and [Buenaposada06]. All publications and works that use this implementations must reference our papers.
-
Implementation of "Boosting Sex Identification Performance" IJCV paper (Baluja-Rowley) matlab toolbox. In this toolbox we have implemented the Baluja and Rowley's IJCV paper (see [Baluja07]). All publications and works that use this toolbox must reference our PAMI 2011 paper [Bekios-Calfa2011].
- SSD-based plane tracking matlab toolbox. In this toolbox we have implemented our Jacobian matrix factorisation for the projective motion model (see [Buenaposada02b] and [Hager98]). To test it out of the box, you should download the chessboard calibration template sequence (320x240 images) and decompress it over the same directory as the matlab toolbox (see the README.txt). All publications and works that use this toolbox must reference our ICPR 2002 paper [Buenaposada02b].
-
Cohn-Kanade facial database apriori classification into the six universal expressions (anger, joy, disgust, fear, sadness, surprise) matlab scripts
The image sequences on the Cohn-Kanade facial database (at least on the public available part) are labeled with the FACS Action Units (AUs) present on the last image of each sequence. There is no direct translation from AUs into one of the universal expressions (most of the time the translation is subjetive). We believe that, in order to make fair comparations beetwen different classification results using the Cohn-Kanade database, it is needed to stablish some sort of benchmark (using the same training sequences and apriori classification).
We provide here two matlab scripts that sets which sequences we use and what apriori classification we made on the image sequences (from the Cohn-Kanade database). All publications and works that use this image sequences must reference our PAA Journal paper, [PAA2008].
Data
- Posters sequence and image templates used in our JRTIP paper. All publications and works that use this image sequences must reference our JRTIP paper: "Speeding-up homography estimation in mobile devices". Pablo Márquez-Neila, Javier López-Alberca, José M. Buenaposada, Luis Baumela. Journal of Real-Time Image Processing, (in press).
- Chess board calibration template sequence and image sequences used in our BMVC 2006 paper. All publications and works that use this image sequences must reference our BMVC 2006 paper: "Efficiently estimating facial expression and illumination in appearance-based tracking". J.M. Buenaposada, E. Muñoz, L. Baumela. 2006. British Machine Vision Conference, BMVC 2006. Edinburgh. September 2006.