IGGG (IFT-GLO-GEL-GIF) Computer Vision + Applied Machine Learning Reading Group
(Winter 2017)
Room PLT-3904 or PLT-1120 (check schedule), Every other Friday at 2:30 PM
  Jean-François Lalonde + Philippe Giguère

What we want are interesting and recent papers from TOP CONFERENCES that have a strong emphasis on computer vision and applied machine learning, preferably papers combining them (no toy problems). Please contact me if you would like to present a paper.

Format of the presentation (approxiate guidelines)
Overall, 30 minutes + lots of interaction with audience. In 60 min, we are out the door.
Quickly discuss relevent previous work, to give a context. Have some critical comments about the paper (why do you think it is good, why you think it is not.) We do not want top-notch slides. Screen capture of the pdf/using the whiteboard/your hands/slides from author are perfectly fine. Think of it as a discussion amongst 10-15 people. Assume that the audience has read the paper, but are hungry for a little explanation.

Audience
PLEASE READ THE PRESENTED PAPER! The goal is to figure out, AS A GROUP, the details and intricacies of the paper, and what is the take home message. If  you are in the audience, we expected you to ask questions or contribute in clarifying certain points that are more obscure to others, including the presenter. We all have different technical backgrounds, let's exploit this. No free riding!

A (non-exhaustive) list of potential conferences:
Journals are also acceptable:


Former presentations

Date
Paper
Presenter
Link to material
July 23rd
Deep learning for detecting robotic grasp, I. Lenz, H. Lee, A. Saxena, IJRR 2014. Ludovic Trottier
http://arxiv.org/pdf/1301.3592.pdf
July 30th
Multi-instance Object Segmentation with Occlusion Handling, Chen et al.  CVPR 2015.
Maxime Tremblay
pdf
WEDNESDAY
August 5th
Geodesic Flow Kernel for Unsupervised Domain Adaptation
Alexandre Lemire-Paquin
pdf
supplement material
journal version (as shown by Brahim)
August 13th
Deep Learning from Temporal Coherence in Video, ICML 2009.
Olivier Gagnon
pdf
(related follow-up provided by JF. Lalonde)
August 20th
NO READING GROUP (DAMAS seminars)


August 27th
Overview of global illumination techniques + following article: Global illumination with radiance regression functions. ACM Transactions on Graphics, 32(4), 1, Ren, P., Wang, J., Gong, M., Lin, S., Tong, X., & Guo, B. (2013). Yannick Hold-Geoffroy
link
September 15, 3 PM
Is object localization for free? – Weakly-supervised learning with convolutional neural networks
Marc-André Gardner
link to pdf
(A precursor is OverFeat, thanks J.F.!)
October 6th, 3 PM
FAB-MAP + FAB-MAP 2.0
Sébastien Michaud