Primary competition visual

Arm UNICEF Disaster Vulnerability Challenge

Helping Malawi
$10 000 USD
Challenge completed over 1 year ago
1187 joined
347 active
Starti
Mar 15, 24
Closei
Jun 23, 24
Reveali
Jun 23, 24
About

Traditional houses in most southern African countries, including Malawi, are either circular or rectangular dwelling units with mud-built walls, supported by bamboo or maize stocks and grass-thatched roofs. In Malawi, the houses usually have one or two rooms and span less than 10 meters, housing an average household size of 4.3 people.

Aerial imagery has been taken over parts of Malawi, and these aerial images have been used to label the data.

In the data download section you will find a document that explains how the data was labelled.

Use of other datasets is prohibited, as this model needs to be applied to remote areas not covered by other datasets.

There are 4 772 images in train and 2 045 images in test.

Each category of dwelling is assigned a number: 1 is Other, 2 is Tin and 3 is Thatch.

Note, there are not many "Other" images but it is important to consider them in your model in the case that this solution is implemented in a place where there are other dwelling types.

Files
Description
Files
This zip file contains both train and test images.
Test resembles Train.csv but without the target-related columns. This is the dataset on which you will apply your model to.
Train contains the target. This is the dataset that you will use to train your model.
Is an example of what your submission file should look like. The order of the rows does not matter, but the names of the "ID" must be correct.
Examples of how the images were labelled.