Microsoft Rice Disease Classification Challenge
Can you identify disease in images of rice grown in Egypt?
Prize
$3 000 USD
Time
Ended 6 months ago
Participants
261 active · 833 enrolled
Helping
Egypt
Intermediate
Computer Vision
Classification
Agriculture
Consider Uploading the Data for Use.
Help · 13 May 2022, 14:19 · 3

The size of the image dataset is quite large (1.3gb) and x2 if one is to upload it into a drive or other resources for use, also considering storage capacities. Zindi should consider uploading large datasets into a single public storage system where participants could easily fetch it. Imagine 100+ participants needing to download and upload such size of data.

Discussion 3 answers

You can directly load data from zindi into colab without spending dat much data... Then copy the file to your drive from colab

# This is something I did a while ago to download Zindi datasets

!git clone "https://github.com/yonas-g/zindi-dataset-downloader.git"

!cp -r zindi-dataset-downloader/* .

!mkdir "./data"

!mkdir "./data/images"

import ZindiDownloader.ZindiDownloader as zindi

auth_token = 'key_here'

downloader = zindi.ZindiDataDownloader(auth_token)

# this can be found by inspecting html element of dataset's webpage. CTRL + f

links = [

{"url": "https://api.zindi.africa/v1/competitions/microsoft-rice-disease-classification-challenge/files/Test.csv", "path": "./data"},

{"url": "https://api.zindi.africa/v1/competitions/microsoft-rice-disease-classification-challenge/files/Images.zip", "path": "./data/images"},

{"url": "https://api.zindi.africa/v1/competitions/microsoft-rice-disease-classification-challenge/files/SampleSubmission.csv", "path": "./data"},

{"url": "https://api.zindi.africa/v1/competitions/microsoft-rice-disease-classification-challenge/files/Train.csv", "path": "./data"}

]

for item in links:

downloader.fetch(item["url"], target_path=item["path"])

15 May 2022, 15:55
Upvotes 2