NTIRE 2018 Super-Resolution Challenge - Track 4: Realistic wild - x4

Organized by jwu - Current server time: Oct. 16, 2018, 8:21 p.m. UTC
Reward $16,000

First phase

Development
Jan. 15, 2018, 11:59 p.m. UTC

End

Competition Ends
March 23, 2018, 11:59 p.m. UTC

NTIRE Workshop and Challenges @ CVPR 2018

 

Example-based Single-Image Super-Resolution Challenge

Important dates

  • 2018.01.10 Release of train data (low-res and high-res images) and validation data (only low-res)
  • 2018.01.15 Validation server online
  • 2018.03.15 Final test data release (only low-res), validation data (high-res) released, validation server closed
  • 2018.03.22 Test high-res results submission deadline
  • 2018.03.24 Fact sheets submission deadline
  • 2018.03.24 Code/executable submission deadline
  • 2018.03.28 Final test results release to the participants (check:  https://competitions.codalab.org/competitions/18026#participate)
  • 2018.04.10 Paper submission deadline for top entries from the challenge
  • 2018.06.18 NTIRE workshop and challenge on example-based single-image super-resolution, results and award ceremony (CVPR 2018, Salt lake city, Utah)

Challenge overview

The 3nd edition of NTIRE: New Trends in Image Restoration and Enhancement workshop will be held June 18, 2018 in conjunction with CVPR 2018 in Salt lake city, Utah.

Image restoration and image enhancement are key computer vision tasks, aiming at the restoration of degraded image content or the filling in of missing information. Recent years have witnessed an increased interest from the vision and graphics communities in these fundamental topics of research. Not only has there been a constantly growing flow of related papers, but also substantial progress has been achieved.

Each step forward eases the use of images by people or computers for the fulfillment of further tasks, with image restoration or enhancement serving as an important frontend. Not surprisingly then, there is an ever growing range of applications in fields such as surveillance, the automotive industry, electronics, remote sensing, or medical image analysis. The emergence and ubiquitous use of mobile and wearable devices offer another fertile ground for additional applications and faster methods.

NTIRE workshop aims to provide an overview of the new trends and advances in those areas. Moreover, it will offer an opportunity for academic and industrial attendees to interact and explore collaborations.

Jointly with NTIRE workshop we have an NTIRE challenge on example-based single-image super-resolution, that is, the task of restoration of rich details (high frequencies) in a high resolution image for a single low resolution input image based on a set of prior examples with low and corresponding high resolution images. The challenge has four tracks.

Track 1: "Classic bicubic" follows the classic / standard settings from the single-image super-resolution literature, that is, the degradation operators are the downscaling with bicubic interpolation (imresize Matlab function) of the ground truth high resolution image.

Track 2: "Realistic mild" adverse conditions - assumes that the degradation operators emulating the image acquisition process from a digital camera (such as blur kernel, decimation, downscaling strategy) can be estimated through training pairs of low and high-resolution images. The parameters of the degradation operators are fixed, while the randomly generated blur kernels and their resulting pixel shifts vary from image to image. The large training set of examples of low and corresponding high resolution images are intended for modeling the low to high image resolution mapping relation.

Track 3: "Realistic difficult" adverse conditions - as above, with larger corruptions.

Track 4: "Realisitic wild" conditions - assumes that the degradation process (emulating the image acquisition process from a digital camera) varies from one image to another to further simulate the real world scenario, that is, the parameters of the degradation operators vary under certain intervals, moreover the randomly generated blur kernels and their resulting pixel shifts vary from image to image.  This setting is the closest to real "wild" conditions.

 

The top ranked participants on each track will be awarded and invited to follow the CVPR submission guide for workshops to describe their solution and to submit to the associated NTIRE workshop at CVPR 2018.

Note that for the final ranking and challenge winners we are weighing more the teams/participants with entries in more than one track.

 

"Classic bicubic" track information (1)

"Classic bicubic" track is meant to facilitate the easy deployment of recent proposed methods for the task for example-based single-image super-resolution. It assumes that the degradation operators are the same as commonly used in the recent super-resolution literature. For obtaining the low res images we use the Matlab function "imresize" with default settings (bicubic interpolation) and the desired downscaling factor: 8. In order to eliminate the boundary effect, we first crop the image boundaries by matlab function so that the image height and width are divisible by 8. The ground truth high resolution RGB image is then downscaled to corresponding low resolution images and used either for training, validation, or testing of the methods.  We are making available a large newly collected dataset -DIV2K- of images with a large diversity of contents.

More details are found on the data section of the competition.

"Realistic" tracks information (2,3,4)

"Realistic" tracks go few steps ahead from the classic bicubic interpolation assumed in most of the recent single-image super-resolution literature and considers that at runtime we know the low resolution input image and a set of (training) pairs of low and corresponding high resolution images. No other information is provided about the degrading operators producing the downscaling images. We employ downscaling factor 4 and define 3 tasks: for mild (track 2) and difficult (track 3) conditions in which the same degradation type is applied in all the images of each track and for wild (track 4) conditions in which the degradation operators are the same within an image space but DIFFERENT from one image to another. This setting is the closest to real "wild" conditions.

Each ground truth high resolution RGB image is downscaled to corresponding low resolution images and processed by several degradation steps for training, validation, and testing of the methods. We are making available a large newly collected dataset -DIV2K- of images with a large diversity of contents.

More details are found on the data section of the competition.

Competitions

To learn more about each competition, to participate in the challenge, and to access the collected DIV2K dataset with DIVerse 2K resolution images everybody is invited to register at the following links, accordingly:

The training data is already made available to the registered participants.

Provided Resources

  • Scripts: With the dataset the organizers will provide scripts to facilitate the reproducibility of the images and performance evaluation results after the validation server is online. More information is provided on the data page.
  • Contact: You can use the forum on the data description page (highly recommended!) or directly contact the challenge organizers by email (Radu.Timofte [at] vision.ee.ethz.ch and Shuhang.Gu [at] vision.ee.ethz.ch and jwu [at] vision.ee.ethz.ch) if you have doubts or any question.

NTIRE Workshop and Challenges @ CVPR 2018

 

Example-based Single-Image Super-Resolution Challenge

 

Evaluation

The evaluation consists from the comparison of the restored high resolution images with the ground truth high resolution images. For this we use the standard Peak Signal To Noise Ratio (PSNR) and, complementary, the Structural Similarity (SSIM) index as often employed in the literature. Implementations are found in most of the image processing toolboxes. For each dataset we report the average results over all the processed images belonging to it.

Since some methods do not provide the restoration of the image boundaries we ignore an image boundary of (6+s) pixels, where s is the scaling factor of the setup. We recommand to the participants also to ignore an image boundary of (6+s) pixels for the provided train data. The boundary size is given for the high resolution image.

Attention for realistic wild setting: For the sake of ruling out the kernel shifting and evaluation efficiency, during the evaluation,  we crop the center part of the image with image size 60x60,  then compare the generated results by shifting them up to 40 pixels for four direction to the groundtruth results, the reported PSNR and SSIM are the best scores achived among the shifted candidates.   The running time may take a while, please be patient and thank you for your understanding.

 

For submitting the results, you need to follow these steps:

  1. process the low resolution input images and keep the same name for the high resolution image results as produced by your method (example: for an input file with name "0834x8.png" the output file should be "0834x8.png") 
    Note that the high resolution images should be saved with lossless compression and should have a size of (s*a,s*b,3) for an input image of size (a,b,3) and scale factor s.
  2. create a ZIP archive containing all the high resolution image results named as above and a readme.txt Note that the archive should not include folders, all the images/files should be in the root of the archive.
  3. the readme.txt file should contain the following lines filled in with the runtime per image (in seconds) of the solution, 1 or 0 accordingly if employs CPU or GPU at runtime, and 1 or 0 if employs extra data for training the models or not.
    runtime per image [s] : 3.31
    CPU[1] / GPU[0] : 1
    Extra Data [1] / No Extra Data [0] : 1
    Other description : Solution based on A+ of Timofte et al. ACCV 2014. We have a Matlab/C++ implementation, and report single core CPU runtime. The method was trained on Train 91 of Yang et al. and BSDS 200 of the Berkeley segmentation dataset. 
     The last part of the file can have any description you want about the code producing the provided results (dependencies, link, scripts, etc.)
    The provided information is very important both during the validation period when different teams can compare their results / solutions but also for establishing the final ranking of the teams and their methods.

New Trends in Image Restoration and Enhancement (NTIRE) challenge on example-based single-image super-resolution @ CVPR 2018

 

Example-based Single-Image Super-Resolution Challenge

These are the official rules (terms and conditions) that guvern how the NTIRE challenge on example-based single-image super-resolution 2018 will operate. This challenge will be simply reffered to as the "challenge" or the "contest" throghout the remaining part of these rules and may be named as "NTIRE" or "DIV2K" benchmark, challenge, or contest, elsewhere (our webpage, our documentation, other publications).

In these rules, "we", "our", and "us" refer to the organizers (Radu Timofte and Shuhang Gu and Jiqing Wu from ETH Zurich, CVL) of NTIRE challenge and "you" and "yourself" refer to an eligible contest participant.

Note that these official rules can change during the contest until the start of the final phase. If at any point during the contest the registered participant considers that can not anymore meet the eligibility criteria or does not agree with the changes in the official terms and conditions then it is the responsability of the participant to send an email to the organizers (Radu.Timofte [at] vision.ee.ethz.ch and Shuhang.Gu [at] vision.ee.ethz.ch and Jiqing.Wu [at] vision.ee.ethz.ch) such that to be removed from all the records. Once the contest is over no change is possible in the status of the registered participants and their entries.

1. Contest description

This is a skill-based contest and chance plays no part in the determination of the winner (s).

The goal of the contest is the restoration of a high resolution clean image with high frequencies (small details) from a low resolution input image and the challenge is called example-based single-image super-resolution.

Focus of the contest: it will be made available a newly collected dataset of at least 1000 high quality images collected from Internet and adapted for the specific needs of the challenge. The images are no larger than 2048 pixels on horizontal or vertical direction and have a large diversity of contents. We will refer to this dataset, its partition, and related materials as DIV2K (DIVerse 2K resolution images dataset). The dataset is divided into training, validation and testing data. We focus on two distinct settings: (track 1) example-based single-image super-resolution where the downscaling degradation operators are the same as in the recent literature (imresize Matlab function with bicubic interpolation) which will easy the deployment of the recent developed methods for our challenge, and (tracks 2-4) example-based single-image super-resolution where the degradation operators are kept hidden in the explicit form, the participant will know only the generated training data and validation data to develop and adapt their methods to achieve the best results on test data. The participants will not have access to the ground truth high resolution images from the test data. For each track the ranking of the participants is according to the performance of their methods on the test data. The participants will provide descriptions of their methods, details on (run)time complexity and (extra) data used for modeling. The winners will be determined according to their entries, the reproducibility of the results and uploaded codes or executables, and the above mentioned criteria as judged by the organizers.

2. Tentative contest schedule

The registered participants will be notified by email if any changes are made to the schedule.

  • 2018.01.10 Development phase, release of train data (low-res and high-res images) and validation data (only low-res)
  • 2018.01.15 Validation server online
  • 2018.03.15 Final test data release (only low-res), validation data (high-res) released, validation server closed
  • 2018.03.22 Test high-res results submission deadline
  • 2018.03.24 Fact sheets submission deadline
  • 2018.03.24 Code / executable submission deadline
  • 2018.03.26 Final test results release to the participants
  • 2018.04.02 Paper submission deadline for top entries from the challenge
  • 2018.06.18 NTIRE workshop and challenge on example-based single-image super-resolution, results and award ceremony (CVPR 2018, Salt lake city, Utah)

3. Eligibility

You are eligible to register and compete in this contest only if you meet all the following requirements:

  • you are an individual or a team of people willing to contribute to the open tasks, who accepts to follow the rules of this contest
  • you are not an NTIRE challenge organizer or an employee of NTIRE challenge organizers
  • you are not involved in any part of the administration and execution of this contest
  • you are not a first-degree relative, partner, household member of an employee or of an organizer of this NTIRE challenge or of a person involved in any part of the administration and execution of this contest

This contest is void wherever it is prohibited by law.

Entries submitted but not qualified to enter the contest, it is considered voluntary and for any entry you submit NTIRE reserves the right to evaluate it for scientific purposes, however under no circumstances will such entries qualify for sponsored prizes. If you are an employee, affiliated with or representant of any of the NTIRE challenge sponsors then you are allowed to enter in the contest and get ranked, however, if you will rank among the winners with eligible entries you will receive only a diploma award and none of the sponsored money, GPU or travel grants.

NOTE: industry and research labs are allowed to submit entries and to compete in both validation phase and final test phase. However, in order to get officially ranked on the final test leaderboard and to be eligible for awards the reproducibility of the results is a must and, therefore, the participants need to make available and submit their codes or executables. All the top entries will be checked for reproducibility and marked accordingly.

We will have 3 categories of entries in the final test ranking:
1) checked with publicly released codes 
2) checked with publicly released executable
3) unchecked (with or without released codes or executables)

 

4. Entry

In order to be eligible for judging an entry must meet all the following requirements:

Entry contents: the participants are required to submit image results and (code or executable). To be eligible for prizes, the top ranking participants should publicly release their codes or executables under a license of their choice, taken among popular OSI-approved licenses (http://opensource.org/licenses) and make their code or executables online accessible for a period of not less than one year following the end of the challenge (applies only for top three ranked participants of the competition). To enter the final ranking the participants will need to fill out a survey (fact sheet) briefly describing their method. All the participants are also invited (not mandatory) to submit a paper for peer-reviewing and publication at the NTIRE Workshop and challenge on image super-resolution (to be held on June 18, 2018, Salt lake city, Utah). To be eligible for prizes, the participants score must improve the baseline performance provided by the challenge organizers.

Use of data provided: all data provided by NTIRE are freely available to the participants from the website of the challenge under license terms provided with the data. The data are available only for open research and educational purposes, within the scope of the challenge. NTIRE and the organizers make no warranties regarding the database, including but not limited to warranties of non-infringement or fitness for a particular purpose. The copyright of the images remains in property of their respective owners. By downloading and making use of the data, you accept full responsibility for using the data. You shall defend and indemnify NTIRE and the organizers, including their employees, Trustees, officers and agents, against any and all claims arising from your use of the data. You agree not to redistribute the data without this notice.

  • Test data: The organizers will use the test data for the final evaluation and ranking of the entries. The ground truth test data will no be made available to the participants during the contest.
  • Training and validation data: The organizers will make available to the participants a training dataset with ground truth images and a validation dataset without ground truth images. The ground truth images for validation dataset will be released at the start of the final phase when the test data without ground truth images will be made available.
  • Post-challenge analyses: the organizers may also perform additional post-challenge analyses using extra-data, but without effect on the challenge ranking.
  • Submission: the entries will be on-line submitted via the CodaLab web platform. During development phase, while the validation server is online, the participants will receive immediate feedback on validation data. The final evaluation will be computed automatically on the test data submissions, but the final scores will be released after the challenge is over.
  • Original work, permissions: In addition, by submitting your entry into this contest you confirm that, to the best of your knowledge: - your entry is your own original work; and - your entry only includes material that you own, or that you have permission to use.

5. Potential use of entry

Other than what is set forth below, we are not claiming any ownership rights to your entry. However, by submitting your entry, you:

Are granting us an irrevocable, worldwide right and license, in exchange for your opportunity to participate in the contest and potential prize awards, for the duration of the protection of the copyrights to:

  1. Use, review, assess, test and otherwise analyze results submitted or produced by your code or executable and other material submitted by you in connection with this contest and any future research or contests by the organizers; and
  2. Feature your entry and all its content in connection with the promotion of this contest in all media (now known or later developed);

Agree to sign any necessary documentation that may be required for us and our designees to make use of the rights you granted above;

Understand and acknowledge that us and other entrants may have developed or commissioned materials similar or identical to your submission and you waive any claims you may have resulting from any similarities to your entry;

Understand that we cannot control the incoming information you will disclose to our representatives or our co-sponsor’s representatives in the course of entering, or what our representatives will remember about your entry. You also understand that we will not restrict work assignments of representatives or our co-sponsor’s representatives who have had access to your entry. By entering this contest, you agree that use of information in our representatives’ or our co-sponsor’s representatives unaided memories in the development or deployment of our products or services does not create liability for us under this agreement or copyright or trade secret law;

Understand that you will not receive any compensation or credit for use of your entry, other than what is described in these official rules.

If you do not want to grant us these rights to your entry, please do not enter this contest.

6. Submission of entries

The participants will follow the instructions on the CodaLab website to submit entries

The participants will be registered as mutually exclusive teams. Each team is allowed to submit only one single final entry. We are not responsible for entries that we do not receive for any reason, or for entries that we receive but do not work properly.

The participants must follow the instructions and the rules. We will automatically disqualify incomplete or invalid entries.

7. Judging the entries

The board of NTIRE will select a panel of judges to judge the entries; all judges will be forbidden to enter the contest and will be experts in causality, statistics, machine learning, computer vision, or a related field, or experts in challenge organization. A list of the judges will be made available upon request. The judges will review all eligible entries received and select winners based upon the prediction scores on test data, factsheet description, complexity and reproducibility. The judges will verify that the winners complied with the rules, including that they documented their method by filling out a fact sheet.

The decisions of these judges are final and binding. The distribution of prizes according to the decisions made by the judges will be made within three (3) months after completion of the last round of the contest. If we do not receive a sufficient number of entries meeting the entry requirements, we may, at our discretion based on the above criteria, not award any or all of the contest prizes below. In the event of a tie between any eligible entries, the tie will be broken by giving preference to the earliest submission, using the time stamp of the submission platform.

8. Prizes and Awards

ETH Zurich, KU Leuven, University of California at Merced, Hong Kong Polytechnic University, Alibaba, NVIDIA, SenseTime, CodeOcean, Huawei, Disney Research, Google and Amazon are the financial sponsors of this contest. There will be economic incentive prizes and travel grants for the winners (based on availability) to boost contest participation; these prizes will not require participants to enter into an IP agreement with any of the sponsors, to disclose algorithms, or to deliver source code to them. The participants affiliated with Alibaba, NVIDIA, SenseTime, CodeOcean, Huawei, Disney Research, Google and Amazon agree to not receive any sponsored money, GPU or travel grant in the case they will be among the winners.

Incentive Prizes for each track competitions (tentative, the prizes depend on attracted funds from the sponsors)

  • 1st place: ?00$ + ?Titan X GPU + award certificate
  • 2nd place: ?00$ + ?Titan X GPU + award certificate
  • 3rd place: ?00$ + award certificate

9. Other Sponsored Events

Publishing papers is optional and will not be a condition to entering the challenge or winning prizes. The top ranking participants are invited to submit a maximum 8-pages paper (CVPR2018 author rules) for peer-reviewing to NTIRE workshop.

The results of the challenge will be published together with NTIRE 2018 workshop paper in the 2018 CVPR Workshops proceedings.

The top ranked participants and participants contributing interesting and novel methods to the challenge will be invited to be co-authors of the challenge report paper which will be published in the 2018 CVPR Workshops proceedings. A detailed description of the ranked solution as well as the reproducibility of the results are a must to be an eligible co-author.

10. Notifications

If there is any change to data, schedule, instructions of participation, or these rules, the registered participants will be notified at the email they provided with the registration.

Within seven days following the determination of winners we will send a notification to the potential winners. If the notification that we send is returned as undeliverable, or you are otherwise unreachable for any reason, we may award the prize to an alternate winner, unless forbidden by applicable law.

The prize such as money, GPU, or travel grant will be delivered to the registered team leader given that the team is not affiliated with any of the sponsors. It's up to the team to share the prize. If this person becomes unavailable for any reason, the prize will be delivered to be the authorized account holder of the e-mail address used to make the winning entry.

If you are a potential winner, we may require you to sign a declaration of eligibility, use, indemnity and liability/publicity release and applicable tax forms. If you are a potential winner and are a minor in your place of residence, and we require that your parent or legal guardian will be designated as the winner, and we may require that they sign a declaration of eligibility, use, indemnity and liability/publicity release on your behalf. If you, (or your parent/legal guardian if applicable), do not sign and return these required forms within the time period listed on the winner notification message, we may disqualify you (or the designated parent/legal guardian) and select an alternate selected winner.

 


The terms and conditions are inspired by and use verbatim text from the `Terms and conditions' of NTIRE 2017 Super-Resolution Challenge and ChaLearn Looking at People Challenges.

NTIRE Workshop and Challenge @ CVPR 2018

 

Organizers

 

The NTIRE challenge on Example-based Single-Image Super-Resolution is organized jointly with the NTIRE 2018 workshop. The results of the challenge will be published at NTIRE 2018 workshop and in the CVPR 2018 Workshops proceedings.

 

Radu Timofte (Radu.Timofte [at] vision.ee.ethz.ch) and Shuhang.Gu [at] vision.ee.ethz.ch and Jiqing Wu (jwu [at] vision.ee.ethz.ch) are the contact persons and direct managers of the NTIRE challenge.

 

More information about NTIRE workshop and challenge organizers is available here: http://www.vision.ee.ethz.ch/ntire18/

Development

Start: Jan. 15, 2018, 11:59 p.m.

Testing

Start: March 15, 2018, 11:59 p.m.

Competition Ends

March 23, 2018, 11:59 p.m.

You must be logged in to participate in competitions.

Sign In