Skip to content

Commit c378dbb

Browse files
authored
Add 'size' details to task descriptions
1 parent 53ad9bc commit c378dbb

File tree

1 file changed

+7
-5
lines changed

1 file changed

+7
-5
lines changed

Diff for: docs/CONTRIBUTING.md

+7-5
Original file line numberDiff line numberDiff line change
@@ -1,8 +1,8 @@
11
# CONTRIBUTING
22
This project is inherently open-source and collaborative. We will structure this respository and project to faciliate volunteer contributions. Please follow the guidelines available below to get started and contribute as efficiently as possible.
33
Just a few quick notes:
4-
- Algorithms should use open-source technologies.
5-
- Dataset used should be open-source or easily available and trackable.
4+
- Algorithms should use open-source technologies.
5+
- Dataset used should be open-source or easily available and trackable.
66

77

88
## Initial data and datasets involved
@@ -21,17 +21,20 @@ Below is a tentative outline of the proposed tasks to estimate the extent and po
2121
- Data acquisition
2222
- List of coordinates for the settlements of interest (formal or informal).
2323
- Create dataset of public satellite imagery containing each settlement
24-
- Contour detection and size estimation
24+
- Contour detection and size estimation (At this stage, a square meter estimate should be enough. However, it is important to keep in mind that ultimately, the most useable size metric is `population`.)
2525
- Validate results visually or with existing location datasets.
2626

2727
### Task I-b (Optional - Not needed if results of Phase I are satisfactory)
2828
- Data augmentation
2929
- Use human input (i.e. Mechanical Turk) to create a training dataset with crowd-sourced contours.
3030
- New iteration of contour / sizing algorithms with added input for accuracy. The additional info and human input will allow the use of more complex algorithms for training.
3131

32+
3233
### Task II - Population Estimation
3334
- Research settlement size estimates at different dates.
34-
- Explore algorithms to estimate settlement population including extrapolation/downscaling of existing datasets
35+
- Explore algorithms to estimate settlement population including extrapolation/downscaling of existing datasets.
36+
- Ultimately, population is most actionable size metric. A few use-cases are listed in [APPLICATIONS](APPLICATIONS.md); among other things, estimations would be used for tracking population displacements and resource allocation.
37+
3538

3639
### Task III - Tracking
3740
- Test algorithms on “live” data, i.e. Planet or DG.
@@ -60,4 +63,3 @@ Usually, we will try and keep issues with tags to orient contributors. Here are
6063

6164
## Maintainer
6265
The maintainer associated to this project is `ericboucher`. If you have any questions, you can contact him at [[email protected]](mailto:[email protected])
63-

0 commit comments

Comments
 (0)