<div dir="ltr"><div><div>Apologies for multiple copies<br>******************************<wbr>****<br><br>Call For Participation:<br>
<br><span class="gmail-il">NTIRE</span>: <span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il">New Trends in <span class="gmail-il">Image</span> Restoration and Enhancement <br><span class="gmail-il">Challenge</span> on Example-based Single-<span class="gmail-il">Image</span> <span class="gmail-il">Super</span>-<span class="gmail-il">Resolution</span><br><br>started February 14th, <span class="gmail-il">2017</span> via CodaLab platform.<br><br><a target="_blank" href="http://www.vision.ee.ethz.ch/ntire17/">http://www.<span class="gmail-m_-6370064308949487862gmail-il">vision</span>.ee.ethz.ch/n<wbr>tire17/</a><br>_____________________<br><br></span><span class="gmail-il">NTIRE</span> organizes two parallel <span class="gmail-il">challenge</span> tracks on <span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il">Example-based Single-<span class="gmail-il">Image</span> <span class="gmail-il">Super</span>-<span class="gmail-il">Resolution</span></span><br><br>=Track 1: <span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il">`bicubic' downscaling (Matlab imresize)  - a commonly employed setup</span> in the literature.<br></div><div>The `bicubic' setup allows for easy deployment of the recent published <span class="gmail-il">super</span>-<span class="gmail-il">resolution</span> methods.<br><br></div>=Track
 2: `unknown' downgrading operators - but with available train data of 
low and high res corresponding images. The `unknown' setup constrains 
the participants to work with the provided train data and to develop 
robust solutions to learn from a limited set of examples.<br><div><br>The <span class="gmail-il">challenge</span> aims to gauge the state-of-the-art, to compare and to promote different solutions with the new DIV2K dataset.<br><br>=DIV2K
 - a newly collected dataset of DIVerse 2K <span class="gmail-il">resolution</span> images 
(2048x[unspecified] pixel images) will be released with the <span class="gmail-il">NTIRE</span> 
<span class="gmail-il">challenge</span>. DIV2K dataset is divided into 80%/10%/10% for 
train/validation/test, resp.<br></div><div>For each track the low res 
images are generated and provided to the registered participants for x2,
 x3, and x4 downscaling factors.<br></div><div><br><span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il">Details about <span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il"><span class="gmail-il">NTIRE</span> <span class="gmail-il">challenge</span></span> and links to its 6 competitions (2 tracks X 3 downscaling factors) can be found below:  </span><br><span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il"><span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il"><a target="_blank" href="http://www.vision.ee.ethz.ch/ntire17/">http://www.<span class="gmail-m_-6370064308949487862gmail-il">vision</span>.ee.ethz.ch/n<wbr>tire17/</a> </span>
<br></span><ul><li><a target="_blank" href="https://competitions.codalab.org/competitions/16303">Track 1: bicubic downscaling x2 competition</a></li><li><a target="_blank" href="https://competitions.codalab.org/competitions/16304">Track 1: bicubic downscaling x3 competition</a></li><li><a target="_blank" href="https://competitions.codalab.org/competitions/16305">Track 1: bicubic downscaling x4 competition</a></li><li><a target="_blank" href="https://competitions.codalab.org/competitions/16306">Track 2: unknown downscaling x2 competition</a></li><li><a target="_blank" href="https://competitions.codalab.org/competitions/16307">Track 2: unknown downscaling x3 competition</a></li><li><a target="_blank" href="https://competitions.codalab.org/competitions/16308">Track 2: unknown downscaling x4 competition</a></li></ul>____________<br><br>
Important Dates <br>____________<br><ul><li><span class="gmail-il">2017</span>.02.14 Release of train data (low-res and high-res images) and validation data (only low-res)</li><li><span class="gmail-il">2017</span>.02.25 Validation server online</li><li><span class="gmail-il">2017</span>.03.20 Final test data release (only low-res), validation data (high-res) released, validation server closed</li><li><span class="gmail-il">2017</span>.03.31 Test high-res results submission deadline</li><li><span class="gmail-il">2017</span>.04.03 Fact sheets submission deadline</li><li><span class="gmail-il">2017</span>.04.03 Code submission deadline</li><li><span class="gmail-il">2017</span>.04.10 Final test results release to the participants</li><li><span class="gmail-il">2017</span>.04.24 Paper submission deadline for top entries from the <span class="gmail-il">challenge</span></li><li><span class="gmail-il">2017</span>.07.21 <span class="gmail-il">NTIRE</span> <span class="gmail-il">workshop</span> and <span class="gmail-il">challenge</span> on example-based 
single-<span class="gmail-il">image</span> <span class="gmail-il">super</span>-<span class="gmail-il">resolution</span>, results and award ceremony (@ <span class="gmail-il">CVPR</span> <span class="gmail-il">2017</span>, 
Honolulu, Hawaii)</li></ul>____________<br><br></div><div></div>The top 
ranking participants will be invited to submit papers to <span class="gmail-il">NTIRE</span> <span class="gmail-il">workshop</span> 
and to be co-authors of the <span class="gmail-il">challenge</span> report paper.<br><br></div>We are looking for sponsors for incentive prizes. <br>Contact: <a href="mailto:radu.timofte@vision.ee.ethz.ch">radu.timofte@vision.ee.ethz.ch</a><br><div>
<br>
For more details please visit: <span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il"><span class="gmail-m_-6370064308949487862gmail-m_927159548823032825gmail-il"><a target="_blank" href="http://www.vision.ee.ethz.ch/ntire17/">http://www.<span class="gmail-m_-6370064308949487862gmail-il">vision</span>.ee.ethz.ch/n<wbr>tire17/</a> </span> </span><br>____________</div></div>