Machine Learning and Object Recognition 2017-2018

When: Thursdays 14h00 - 15h30
Where: ENSIMAG, room D109/D111
Who: Lectures by Cordelia Schmid and Jakob Verbeek.
Grading: Final grades are determined as follows
- 50% written exam
- 50% quizes on the presented papers
- Paper presentation, grade replaces worst grade among quizes
Paper presentations:
- each paper is presented by two or three students
- presentations last exactly 20 minutes , time yours in advance!
- Papers are assigned among students on first-serve basis

October 5, D109

Cordelia Schmid
Student presentation: Distinctive image features from scale-invariant keypoints
Florent Chastagner, Anshul Paigwar, Spencer Raisanen
Student presentation: Deep image retrieval: Learning global representations for image search
Manuel Carmago, Pauline Olivier

Cordelia Schmid: Bag-of-features for retrieval [slides] and classification [slides]
Student presentation: Faster R-CNN: Towards real-time object detection with region proposal networks
Mujtaba Asif, Ismail Harrando, Ahmad Khan

Jakob Verbeek
Student presentation: Densely Connected Convolutional Networks
Michel Aractingi, Nieves Crasto and Khoder Jneid
Student presentation: Deep Learning on Point Sets for 3D Classification and Segmentation
Olof Ekborg and Dominik Schreiber

Cordelia Schmid: action recognition [intro_slides], [main_slides] + optical flow [slides]

Jakob Verbeek: Quiz to be handed in by 11/1/2018 in plain text email to jakob.verbeek@inria.fr
Student presentation: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation Amela Fejza, Angan Mitra
Student presentation: Flownet 2.0: Evolution of optical flow estimation with deep networks
Danilo.Cazzolla, Paolo Iannino, Moreno La Quatra
Student presentation: Learning a Predictable and Generative Vector Representation for Objects
Maria Garkavenko, Ekaterina Iakovleva, Vitalii Emelianov

Cordelia Schmid
Student presentation: Two-Stream Convolutional Networks for Action Recognition in Videos
Artem Betlei, Ivan Koval, Clement Perny
Student presentation: Action Tubelet Detector for Spatio-Temporal Action Localization
Brendan Guerot, Erick Santillan-Perez