Phase I: Please access the following forms and fill in the requisite details.
- Data Contributors (Data owners who wish to contribute their data to CRDDC) -- https://forms.gle/m2BsijezYsCFJj5R7
- Information Providers (Data owners who wish to participate by providing only the information related to their dataset, keeping the data private) -- https://forms.gle/sh3jE8uG9b9cSvk39
- Data Recommenders (Researchers/Academicians/Freelancers.... who wants to recommend the inclusion of others' datasets to CRDDC) - https://forms.gle/q2U5xVBbrxHRWd1g8
Phase II: Participants need to provide a link to their proposed dataset for verification.
Please upload output files for models targeting Japan, India, Norway, United States,
and Overall (6 countries) for the respective leaderboards! Average of all leaderboards will be used for evaluation. Submission link can be accessed after you log in.
The task is described in Overview section. The submission file is required to follow the following guidelines:
For each image in the test dataset, your algorithm needs to predict a list of labels, and the corresponding bounding boxes. The output is expected to contain the following two columns:
- ImageId: the id of the test image, for example, India_00001
- PredictionString: the prediction string should be space-delimited of 5 integers. For example, 2 240 170 260 240 means it's label 2, with a bounding box of coordinates (x_min, y_min, x_max, y_max). We accept up to 5 predictions. For example, if you submit 3 42 24 170 186 1 292 28 430 198 4 168 24 292 190 5 299 238 443 374 2 160 195 294 357 6 1 224 135 356 which contains 6 bounding boxes, we will only take the first 5 into consideration.
Report and Source Code submissionPlease fill in the requisite details using following forms.
- Report and Source Code upload -- https://forms.gle/pLF2WNN8TcncnerS8
- Information for Leaderboard 1 (overall (6 countries)) -- https://forms.gle/RqQZVLKdtXNhQ3t78
- Information for Leaderboard 2 (Models for India) -- https://forms.gle/R89VDescTWg1UQDX9
- Information for Leaderboard 3 (Models for Japan) -- https://forms.gle/9hngNxZ1Xxg47J2w6
- Information for Leaderboard 4 (Models for Norway) -- https://forms.gle/ot5abcd5BJkV1uGz8
- Information for Leaderboard 5 (Models for United States) -- https://forms.gle/XJsVmJJoQcLsc3NV9
- Feedback -- https://forms.gle/oYFSUPFFdCQnu8bs5
- After the competition phase is completed, a link for submitting the accompanying academic paper will be provided to the top 10 participants as ranked by the public/private leaderboard weighting and the report/Source-code submission described above.
- Peer reviewers will review the academic papers.
- The papers are expected to conform to the IEEE 2-column format set by the conference, which can be found at Big Data CFP.
- If you have questions, please feel free to contact the lead organizer of the dataset competition.
Contents in technical paper/report (Required):
- Explanation of your method, with complete details of technique used (for instance, ensemble learning or data augmentation etc.)
- Evaluation of your method (you can use results obtained on the site of road damage detection challenge to compare)
- Detailed evaluation of your results (inference speed, model size for final deployment, etc.)
- Code and trained model link
Contents in technical paper (Optional but preferable):
- Error Analysis: Examples of failed attempts, efforts that did not go well.
- Advantages and limitations of using your proposed solution.
Source code will also be required to be submitted, either through a publicly available repository on a Git-based version control hosting service such as GitHub or BitBucket for the final evaluation. All source codes are expected to be released as open-source software, utilizing some generally accepted licensing such as Apache License 2.0, GNU General Public License, MIT license, or others of similar acceptance by the Open Source Initiative.