Primary competition visual

INEGI UN-GGIM Human Settlement Detection Challenge by ITU

Helping Mexico
1 000 CHF
Completed (over 1 year ago)
Classification
208 joined
84 active
Starti
Sep 02, 24
Closei
Sep 28, 24
Reveali
Oct 10, 24
User avatar
marching_learning
Nostalgic Mathematics
IS THE TRAINING LABELS AVAILABLE NOW ?
Data Ā· 4 Sep 2024, 08:17 Ā· 2

Hello everyone. Is the training labels now available ?

PLEASE @Zindi @Amy_Bray @meganomaly, RESPOND. This is a very SHORT competition with 25 DAYS REMAINING and WE HAVE ALREADY LOST 2 DAYS (AND POTENTIALLY 20 SUBMISSIONS). We just want to compete and have fun. But this platform is getting really frustrating since we don't have quick responses like in many other ongoing competitons :

-Smart Energy Supply Scheduling for Green Telecom Challenge by ITU : we struggled to get the scoring function script, we went blindly for roughly 2 weeks.

-Microsoft Learn Location Mention Recognition Challenge : We reports inconsistencies in labels that can make the top LB a lottery (no clear answer yet). Personally I took a break on this one.

-GeoAI Amazon Basin Secret Runway Detection Challenge by ITU: No active participant for one month now, no sample submission file, no clear explanation.

-INEGI-GCIM Vegetation Mapping Challenge by ITU : The LB score is unclear, we don' know what is the evaluation metric (no answer yet)

-INEGI-GCIM Human Settlement Detection Challenge by ITU : The training labels are missing (still waiting...).

Discussion 2 answers
User avatar
Amy_Bray
Zindi

Thank you for sharing your feedback and highlighting the issues you’ve encountered. We understand the frustrations some participants have been feeling, and we want to take this opportunity to address each of the concerns you’ve raised. Our goal is to ensure that every competition runs smoothly, and we are constantly working to improve the experience for everyone involved. Below, we’ve provided responses to each of the challenges mentioned in your post.

Smart Energy Supply Scheduling for Green Telecom Challenge by ITU We acknowledge the delay in providing the scoring function script, and we sincerely apologize for the inconvenience caused. This was a new error metric for both Zindi and the team, and we worked diligently to resolve it. Normally, we would consider extending the deadline, but due to the tight timeline with the top winner being invited to COP29 on November 11 in Azerbaijan, an extension is not possible. The error has now been fixed, and we appreciate your patience.

GeoAI Amazon Basin Secret Runway Detection Challenge by ITU We are actively working with the client to introduce a new metric for this challenge. As soon as everything is finalized, we will notify participants through a post on the discussion board. Your feedback is important, and we are committed to improving the experience.

INEGI-GCIM Vegetation Mapping Challenge by ITU We apologize for the confusion regarding the leaderboard (LB) score. There was a typo in the original post, and the correct evaluation metric is log loss. The evaluation page has now been updated with the correct information.

INEGI-GCIM Human Settlement Detection Challenge by ITU We sincerely apologize for the delay in providing the correct data files. The missing training labels "are in the .h5 files that" have now been uploaded, and you can find them on the data page or through this link: https://drive.google.com/drive/folders/10KqkaOAGKKdYK6qrFce3chwW7sIeyDWS?usp=sharing

Microsoft Learn Location Mention Recognition Challenge Thank you for raising concerns about label inconsistencies. We are investigating the issue and will post a detailed update on this specific challenge page shortly.

We understand your frustration and want to assure you that the team is doing everything possible to address the issues as quickly as we can. We appreciate your feedback and value your participation in these competitions.

5 Sep 2024, 09:08
Upvotes 1
User avatar
marching_learning
Nostalgic Mathematics

Thank you very much @Amy_Bray 👏👏👏