[4eyes] FW: Objects365 Dataset and Competition Released!

[Matthew Turk]

FYI – looks like an interesting dataset to know about.
 
From: Biendata <competition at biendata.com> 
Sent: Friday, May 10, 2019 2:14 AM
To: mturk at cs.ucsb.edu
Subject: Objects365 Dataset and Competition Released!
 






  <https://gallery.mailchimp.com/10bd801a2cbd442045e3c3ee3/images/8d4ef596-e891-4c9e-9a70-8285f0c70fab.jpg> 






Objects 365 Challenge Track
 
The Objects365 is a brand new dataset, designed to spur object detection research with a focus on diverse objects in the Wild. Objects365 has 365 object classes annotated on 638K images, totally with more than 10 million bounding boxes in the training set. Thus the annotations cover common objects occurring in all kinds of scene categories. Challenge for Objects365 is proposed to have two tracks, which are Full Track and Tiny Track.
 


 <https://Biendata.us18.list-manage.com/track/click?u=10bd801a2cbd442045e3c3ee3&id=3d7c47d920&e=d729d2bbcd> Objects365 Challenge Track 
 



  <https://gallery.mailchimp.com/10bd801a2cbd442045e3c3ee3/images/35eede4c-0218-4220-a296-4d9465cbcc93.jpg> 

Objects365 Full Track 

The goal of Full Track is to explore the upper-bound performance of object detection systems based on 365 classes and 600K+ training images. 30K images are used for validation and 100K additional images are used for testing. To evaluate the performance of object detection, the evaluation criteria for COCO (IOU from 0.5 to 0.95) benchmark will be adopted.
 

	
 



  <https://gallery.mailchimp.com/10bd801a2cbd442045e3c3ee3/images/38f45f87-c27b-4af1-ad6c-4f3eab751911.jpg> 

Objects365 Tiny Track 

The Tiny Track is to lower the entry threshold, accelerate the algorithm iteration speed, and study the long tail category detection problem. >From the Objects365 dataset, 65 categories are selected, and contestants can train their models by using 10K training data.
 

	
 

	


Prizes 

The total cash prize is $20,000.
Gold Medal (1st Place) for the Objects 365 Full Track: 10,000 USD.
Gold Medal (1st Place) for the Objects 365 Tiny Track: 10,000 USD.
 
 

	
 



Timeline:

April 22, 2019 : The challenge website is officially released, registration opens.
April 29, 2019 : Release the training set and verification set. Participants can submit their forecasting results on the validation data set.
May 10, 2019(23:59p.m PST) : Release the test data. Participants can submit their forecasting results on the test data set (no more than 5 times during the competition).  
June 10, 2019 (23:59p.m PST) :  Submission deadline for both the validation set and the test set. Challenge ends. 
June 17, 2019 : Announcement of final winners at CVPR conference. The champions will be asked to submit a document with at least 4 pages which explains their models. They will be asked to hold a Q & A session at the CVPR workshop.
Note：All times are PST.
 


 <https://Biendata.us18.list-manage.com/track/click?u=10bd801a2cbd442045e3c3ee3&id=fa5f2048b0&e=d729d2bbcd> Objects365 Challenge Track 
 

	
 



Contact Us

 If you have any questions or concerns, please feel free to contact us
Biendata: support at biendata.com <mailto:support at biendata.com> 
Objects 365 : info at objects365.org <mailto:info at objects365.org> 
OR
Join in  <https://Biendata.us18.list-manage.com/track/click?u=10bd801a2cbd442045e3c3ee3&id=811a1a1e7b&e=d729d2bbcd> Discussion Board to find more!

* We also have a WeChat group for competitions, please do not hesitate to join in if you have WeChat. WeChat ID : shujujingsai
 




	


Copyright at 2019 Biendata.com, All rights reserved.
Our Website address is :
Biendata.com








This email was sent to mturk at cs.ucsb.edu <mailto:mturk at cs.ucsb.edu>  
 <https://Biendata.us18.list-manage.com/about?u=10bd801a2cbd442045e3c3ee3&id=d40daf5090&e=d729d2bbcd&c=7bf6e36002> why did I get this?    unsubscribe from this list <https://Biendata.us18.list-manage.com/unsubscribe?u=10bd801a2cbd442045e3c3ee3&id=d40daf5090&e=d729d2bbcd&c=7bf6e36002>     update subscription preferences <https://Biendata.us18.list-manage.com/profile?u=10bd801a2cbd442045e3c3ee3&id=d40daf5090&e=d729d2bbcd>  
Biendata · Floor 7, Building AB · Yuan Yang Guang Hua Guo Ji · Beijing, Beijing 100010 · China 
  <https://Biendata.us18.list-manage.com/track/open.php?u=10bd801a2cbd442045e3c3ee3&id=7bf6e36002&e=d729d2bbcd> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cs.ucsb.edu/pipermail/ilab-users/attachments/20190510/a849d9f2/attachment-0001.html>