Machine Learning and Category Representation 2013-2014

When: Fridays between 9h45-12h45.
Where: ENSIMAG, rooms change, see lectures.
Who: Lectures by Cordelia Schmid and Jakob Verbeek.
Grading: Final grades are determined as follows
- 50% written exam,
- 25% paper presentation (one per student),
- 25% quizes on the presented papers.

Schedule is subject to changes, please check regularly.

Class 1: November 15, 2013, Room H104

Cordelia Schmid: Local invariant features slides1 slides2

Class 2: November 22, 2013, Room H104

Cordelia Schmid: Instance-level and category-level recognition slides1 slides2
Students selects papers they will present.

Class 3: November 29, 2013, Room H201

Cordelia Schmid: object category localization and action recognition slides1 slides2

Class 4: December 6, 2013, Room H201

Jakob Verbeek: Clustering with k-means, and EM for mixture of Gaussians, slides.
Student presentation 1: Segmentation as selective search for object recognition, Elie Cattan, slides.
Student presentation 2: Actom sequence models for efficient action dectection, Benoit Masse, slides.

Class 5: December 13, 2013, Room H201

Matthijs Douze: Fisher vector image representation, slides.
Student presentation 3: Action Recognition with Improved Trajectories, Hoa Nguyen, slides.

Class 6: December 20, 2013, Room H201

Jakob Verbeek: classification methods slides.
Student presentation 4: Segmentation driven object detection with Fisher vectors, Camille Brasseur, slides.
Student presentation 5: Metric Learning for Large Scale Image Classification: Generalizing to New Classes at Near-Zero Cost, Ahmad-Mustofa Hadi, slides.

Exam

9-12 am, January 24, 2014, H202.
prepare from lecture slides, presented papers, and last year's exam.
You can bring printouts of the lecture slides, and the slides of the student presentations, nothing else !