So, in this challenge, your task is to detect the icebergs as soon as possible for alerting the crew to take action with limited computation resources & time.
Our next puzzle is a subject of interest to both polar bears and Leonardo Di Caprio! ICEBERGS!
From the famous sinking of the Titanic ship to the loss of habits of penguins, iceberg detection is of great interest! With the rising temperatures due to global warming, more chunks of ice are finding their way into a water body. The heavy fogs make identifying icebergs even more tricky! Not identifying these icebergs in time can cause major damage to ships and even sink them!
In our next puzzle, your task is to detect icebergs using video input to avoid any damage! Your task is to return the segmentation for the icebergs in a grayscale image.
💪 Getting Started
So, in this challenge, your task is to detect the icebergs as soon as possible for alerting the crew to take action. Only 1 CPU & 2 GB Ram will be available during evaluation!
Use our Getting Started Notebook available here.
In this dataset, for example, train.zip. Each video will represent the icebergs such as -
And your task will be to return the segmentation for the icebergs in a grayscale video where the pixels are only 0 ( Water ) and 255 ( Iceberg ).
- Make sure the video has a total of 23 frames.
- The video output is a lossless format ( i.e no compression ). To learn about how to output video with no compression, try out the starter notebook.
Following files are available in the resources section:
- train.zip - [ 2000 samples ] Used for training.
- test.zip - [ 500 samples ] Used for evaluation, does not include labels
- This challenge accepts the notebook as a submission.
- During the evaluation, the Define preprocessing code 💻 and Prediction phase 🔎 parts notebook will be run, so please make sure it runs without any errors before submitting.
- The notebook follows a particular format, please stick to it.
- Do not delete the header of the cells in the notebook.
This challenge is a notebook execution submission means that your submission notebook will the executed in a cloud server with resources of -
- 1 CPU
- 2 GB Ram
The timeout for submission is over 15 minutes. So make sure your notebook predictions section runs with CPU and in under 10 minutes.
And Let us surely know in Discussion Section if you have any Doubts or Issues :)
🖊 Evaluation Criteria
During the evaluation, the Dice Coefficient will be used to test the efficiency of the model.
- 💪 Challenge Page: https://www.aicrowd.com/challenges/icebergs-detection
- 🗣️ Discussion Forum: https://www.aicrowd.com/challenges/icebergs-detection/discussion
- 🏆 Leaderboard: https://www.aicrowd.com/challenges/icebergs-detection/leaderboards