Machine Learning and Category Representation 2014-2015

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,
- 25% quizes on the presented papers.
Paper presentations:
- each student presents once,
- each paper is presented by a team of two students,
- presentations last for 15 minutes, time yours in advance!

Class 1: November 21, 2014, Room H203

Cordelia Schmid: Local invariant features [slides]
Jakob Verbeek: Clustering with k-means and mixture of Gaussians [slides]
Students selects papers they will present.

Class 2: November 28, 2014, Room H203

Jakob Verbeek: Generative classification [slides]
Cordelia Schmid: Efficient visual search of local features [slides]

Class 3: December 5, 2014, Room H206

Cordelia Schmid: Bag-of-features for category classification [slides], Category localization [slides]
Student presentation: Distinctive image features from scale-invariant keypoints, Theophile Plancke and Mathilde Andre. [slides]
Student presentation: Structured Forests for Fast Edge Detection, Mohammad Shaker, Vladyslav Sydorov, and Nestor Bohorquez.[slides]

Class 4: December 12, 2014, Room H203

Jakob Verbeek: Discriminative classification [slides]
Student presentation: Beyond bags of features: spatial pyramid matching for recognizing natural scene categories , Ayush Tewari, Guillaume Cordonnier, and Andreea-Valeria Vesa. [slides]
Student presentation: Efficient Subwindow Search: a Branch and Bound Framework for Object Localization, Nishant Jain, Jocelyn Meyron, and Alexandre Bleron. [slides]

Class 5: December 19, 2014, Room H203

Cordelia Schmid: Action recognition [slides]
Student presentation: Segmentation as selective search for object recognition, Gregory Cano, Carole Plasson, and Jeremy Wambecke. [slides]
Student presentation: Rich feature hierarchies for accurate object detection and semantic segmentation, Johanna Delanoy, Paul Cottalorda, and Jacques Kaiser. [slides]

Class 6: January 9, 2015, Room H203

Jakob Verbeek: Fisher vector image representation [slides]
Student presentation: Action Recognition with Improved Trajectories, Yassine Bargach, Yassine Karim, and Abdessalam El Jai. [slides]
Student presentation: Fisher vector faces in the wild , Sabina Kruk, Maxime Portaz. [slides]

Exam

9:00-12:00 am, January 30, 2015, room H105.
prepare from lecture slides, presented papers, and last year's exam.
You can bring printouts of the research paper presented in the lectures, but nothing else.