The 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) is over. CVPR’2017 is the premier annual computer vision event comprising the main conference and several co-located workshops, the best meeting point for students, academics and industry researchers in this area. This year Deep Learning and Artificial Intelligence were the focus of the conference, an this was the reason why BSC-CNS has been present with contributions in two workshops:
- Disentangling Motion, Foreground and Background Features in Videos: Our model and source code are publicly available at this https URL The paper that contains the detailed contribution can be downloaded from arxiv repository.
- SalGAN: Visual Saliency Prediction with Generative Adversarial Networks: Our model and source code are publicly available at this https URL . The shorter extended abstract presented as spotlight in the CVPR 2017 Scene Understanding Workshop (SUNw) is available here.
It was an awesome experience for BSC, where we really learned a lot in this area and found many new applications of High Performance Computing.
We also met many people, in order to explain them what BSC-CNS is. Especially researchers who have some kind of bond with Barcelona, to show them the opportunities offered by the BSC-CNS. Here some of them:
Amaia Salvador, Àngels Ratés , José M. Álvarez VAREZ, Francesc Moreno-Noguer, Cristian Canton, Àgata Lapedriza, Jordi Pont-Tuset, Laura Leal-Taixe and Ricardo Martín. Hope to see all of you at BSC-CNS soon!
Also we meet many researchers from industry. We could discuss with outstanding researchers working on issues related to ours. We invited them to visit our center in Barcelona. Here pictures with some of them from Google:
(*) Let me highline a noticeably different from the conferences I used to attend until now in Supercomputing area. With regard to the disclosure of research contributions I think that Supercomputing community should thing about it: most of the papers presented in the main conference and workshops are published on arXiv even before conference peer review. As Yann LeCun comment in Facebook, “Sometimes, by the time a “hot” paper gets presented at a conference, it already has over 50 citations on Google Scholar”. Interesting, right?