Crowdsourcing in ML Boot Camp. We consider mIOU without pictures for a new task from Odnoklassniki
Hello! Hot summer. The organizers of the IT championships sat a lot in the sun, burned and grabbed a blow, but the main thing was to assemble a new task for the next (already the ninth) contest on the ML Boot Camp platform. Championship will be held online within a month.
Now more than 10 thousand specialists are registered on the platform. It often happens that the tasks that appear on it are not all in the teeth (including the Kaggle gods). What are we doing this for? It is necessary to develop and try, and on real data, and not synthetic bulshite. Victories will come with time.
The task that we want to offer you to solve in this competition is different from all previous ones. The subject of the task is the detection of objects in images. Task formulation implies that there will be pictures in the data set, but, funny, they are not. And this is not petabytes of data. And not even gigabytes.
The topic is popular and, in fact, very important. In the Odnoklassniki social network, as in other products, there is a task to filter content. Imagine that your child sees the wrong picture - his world will change forever.
')
Two years ago, Alexey Sennikov wrote an article about how Odnoklassniki attract their users to solve this problem. In short, in the article, Alexey talked about the Odnoklassniki moderator gaming application, in which social network users classify the images into good and bad, getting various buns for it, and also shared a way to solve the problem of increasing DAU (Daily Active Users) applications using algorithms machine learning.
By the way, DAU now has 40,000 users of the application, and more than 1'000'000 solutions per day. Not bad, huh?
In the Odnoklassniki social network there is a platform for marking data. On it, users were given the task to select a given object in a photo with a rectangle. According to the answers of the people, the true position of the object must be restored.
At the entrance you will receive crowdsourcing markup and data in the form of all your favorite labels. In response, you will need to send a simple prediction file.
Dataset and Baseline will be published on the day the competition starts at ML Boot Camp.
To evaluate the solution will be used metric mIOU (mean intersection over union). If you have not yet encountered this metric, we recommend reading the article about it.
Championship will go in online format. We start on June 27 at 19:00, close the submissions on July 29 at 12:00. The results are also summed up on July 29 at 13:00.
All registered users on the platform will receive a reminder to the mail. Sign up or subscribe to the @mrgchamps channel . Along the way, log in to the community of members ( chat @mlbootcamp for 1,400 people) in Telegram to be exactly the topic of everything that happens.
Where do without them. First, the best participants will receive:
1st place: MacBook Pro 13 ', 2-core processor, 256 GB SSD, 16 RAM.
2nd place: iPhone XS Max, 256 GB.
3 - 6 place: Apple Watch Series 3 42mm or Samsung Gear S3 Frontier to choose from.
7 - 10 place: Western Digital My Passport 4 TB.
Secondly, T-shirts with a cool print will get the top 30% of users from the total number of participants.
For a good start in the championship, we recommend reading the articles:
Come, participate, learn and win!
Now more than 10 thousand specialists are registered on the platform. It often happens that the tasks that appear on it are not all in the teeth (including the Kaggle gods). What are we doing this for? It is necessary to develop and try, and on real data, and not synthetic bulshite. Victories will come with time.
The task that we want to offer you to solve in this competition is different from all previous ones. The subject of the task is the detection of objects in images. Task formulation implies that there will be pictures in the data set, but, funny, they are not. And this is not petabytes of data. And not even gigabytes.
The topic is popular and, in fact, very important. In the Odnoklassniki social network, as in other products, there is a task to filter content. Imagine that your child sees the wrong picture - his world will change forever.
')
Two years ago, Alexey Sennikov wrote an article about how Odnoklassniki attract their users to solve this problem. In short, in the article, Alexey talked about the Odnoklassniki moderator gaming application, in which social network users classify the images into good and bad, getting various buns for it, and also shared a way to solve the problem of increasing DAU (Daily Active Users) applications using algorithms machine learning.
By the way, DAU now has 40,000 users of the application, and more than 1'000'000 solutions per day. Not bad, huh?
The challenge of the new contest
In the Odnoklassniki social network there is a platform for marking data. On it, users were given the task to select a given object in a photo with a rectangle. According to the answers of the people, the true position of the object must be restored.
At the entrance you will receive crowdsourcing markup and data in the form of all your favorite labels. In response, you will need to send a simple prediction file.
Dataset and Baseline will be published on the day the competition starts at ML Boot Camp.
To evaluate the solution will be used metric mIOU (mean intersection over union). If you have not yet encountered this metric, we recommend reading the article about it.
schedule
Championship will go in online format. We start on June 27 at 19:00, close the submissions on July 29 at 12:00. The results are also summed up on July 29 at 13:00.
All registered users on the platform will receive a reminder to the mail. Sign up or subscribe to the @mrgchamps channel . Along the way, log in to the community of members ( chat @mlbootcamp for 1,400 people) in Telegram to be exactly the topic of everything that happens.
Presents
Where do without them. First, the best participants will receive:
1st place: MacBook Pro 13 ', 2-core processor, 256 GB SSD, 16 RAM.
2nd place: iPhone XS Max, 256 GB.
3 - 6 place: Apple Watch Series 3 42mm or Samsung Gear S3 Frontier to choose from.
7 - 10 place: Western Digital My Passport 4 TB.
Secondly, T-shirts with a cool print will get the top 30% of users from the total number of participants.
Literature
For a good start in the championship, we recommend reading the articles:
- Grouper: Optimizing Crowdsourced Face Annotations
- Maximum Likelihood Estimation of Error-Rates Using the EM Algorithm
Come, participate, learn and win!
Source: https://habr.com/ru/post/456760/
All Articles