Recognizing Families In the Wild Data Challenge (5th Edition) in conjunction with FG 2021 Forum

Go back to competition Back to thread list Post in this thread

> Deadline of 4th Recognizing Families In the Wild: An IEEE 2020 FG Challenge extended to 2/1/20

Greetings Martin Knoche:

Due to the small window between the end of the challenges and the paper due date, we have decided to extend paper submissions (along with supplemental PDF, movie demos, and code). 4th Recognizing Families In the Wild: An IEEE 2020 FG Challenge, all tracks, will be open for submissions until the end of the day on 1 February 2020 (time-zone, AoE).

We hope this helps and do let us know of any questions or requests.

Sincerely,

2020 RFIW Committee

Posted by: jvision @ Jan. 29, 2020, 2:46 a.m.

From the introduction, we konw that challenge call for challenge papers, general papers or Brave New Ideas, but should we submit three papers (i.e., three challenge papers) for three tasks or one paper for three tasks?
Thanks !

Posted by: DeepBlueAI @ Jan. 29, 2020, 4:26 a.m.

From the introduction, we konw that challenge call for challenge papers, general papers or Brave New Ideas, but should we submit three papers (i.e., three challenge papers) for three tasks or one paper for three tasks?
Thanks !

Posted by: DeepBlueAI @ Jan. 29, 2020, 4:26 a.m.
That is totally up to you. Furthermore, that would be dependent on the extent each task was done. In other words, if the work seems to be a paper in itself, then perhaps 2 or more submissions.

However, I would expect 90% of the time it would be better to get all content in a single, well-written document (i.e., 1 dense, nicely presented paper versus several of lower quality).

For this, either notebook-style works (i.e., more organized around the task, as exemplified in the write-up shared in an earlier forum). Otherwise, it could be organized as more of a story (i.e., conventional research paper, which includes all settings and results in an experimental section).

Just a few notes. At a minimum, the following should be included:
-- abstract
-- overview/introduction
-- methodology used, along with data and related works that inspired this
- Each task and method used should be explicitly described, if more than 1 task per paper
-- analysis of results
- Same if more than 1 task per paper
-- conclusions/ next steps (if any)

The sections do not have to be named as listed above (these are not requirements, as much as content that should be contained within a paper in some fashion). Nonetheless, if in doubt, feel free to have paper organized like the above or using a unique flavor or following another convention (but includes the works)

Does this make sense/ help?

Posted by: jvision @ Jan. 29, 2020, 5:50 p.m.

Yes, Thanks for your reply !

Posted by: DeepBlueAI @ Jan. 30, 2020, 7:52 a.m.
Post in this thread