Scripts I wrote during my master
-
For the probabilistic graphical models course (F. bach)
- K-Means (Done)
- Gaussian Mixture Model (just to commit)
- Hidden Markov Chain (Done)
- Conditional Random Fields (TODO)
-
For the computer vision & object recognition course (J. Sivic, J. Ponce, C. Schmid)
- Project: multiview CCA for caption generation using CNN (ImageNet pretrained) and Word2Vec embedding (TODO)
-
For random matrices course (TODO)
-
Semi-circle law
-
Marcenko Pastur distribution and first order perturbation
-
TP1 (eigein values of large covariance matrices estimation)
-
TP2 (spectral clustering)
-
For the stochastic image analysis course (A. Desolneux)
-
n-grams for text generation
-
binary image segmentation via MRFs (simulated annealing/MCMC for inference)
-
For the kernel methods course (J. Mairal)
-
code of our SVM (ranked 4th on the kaggle)