ICLR Workshop Challenge #2: Radiant Earth Computer Vision for Crop Detection from Satellite Imagery
$5,000 USD
Identify crop type using satellite imagery, and win a trip to present your work at ICLR 2020 in Addis Ababa.
442 data scientists enrolled, 100 on the leaderboard
3 February—29 March
Getting in the top20 with no effort.
published 3 Mar 2020, 16:26

https://colab.research.google.com/drive/1DPizsNT7GUK776TRDmk5rZVMsB1kJY5H#scrollTo=533Q-XzAd-5Q

Johno's notebook which was his starter notebook shared earlier will get you a 1.20 score and in the top20 with little tweaking on the hyperparameters.

Beginners in Computer Vision like me can try it :)

Neat! I guess that's now significant information because of the current update of the rules..

Quick one though.... what reference is "SampleSub.csv" in ss = pd.read_csv('SampleSub.csv') pointing to?

The submission file that is given in the data section of the challenge. It's the file format you need to submit to the platform for it to be a valid submission. He's just reading the original submission file and replacing the target columns with the predictions he generated.

Thank you so much for sharing this.

Hi, please is there a way of generating coordinates for the Sample Sub.csv file? Would have loved to plot these points and visualize them in Google Earth Engine