GeoAI Challenge

Everything happens somewhere – applying machine learning to geospatial analysis

We are calling for GeoAI problem statements to launch the AI for Good GeoAI Challenge.

Geospatial AI (or GeoAI for short) is the discipline that uses AI to analyze data sets which include a spatial (location) component, i.e., a component that can be located by a coordinate system. Most data sets have location coordinates.

We are calling for GeoAI problem statements to launch the AI for Good GeoAI Challenge.

Geospatial AI (or GeoAI for short) is the discipline that uses AI to analyze data sets which include a spatial (location) component, i.e., a component that can be located by a coordinate system. Most data sets have location coordinates.

Related sessions
15 March 2022
15:00 - 17:00 CET, Geneva
Maria Antonia Brovelli (Politecnico di Milano), Vasil Yordanov (‘Vasil Levski’ National Military University)
29 March 2022
15:00 - 17:00 CEST, Geneva
Maria Antonia Brovelli (Politecnico di Milano), Vasil Yordanov (‘Vasil Levski’ National Military University)

GeoAI Challenge Timeline

Now – March 2022

April – July 2022

August – October 2022

Curation Phase

Competition Phase

Evaluation phase

GeoAI Challenge Timeline

Now – March 2022

Curation Phase

April – July 2022

Competition Phase

August – October 2022

Evaluation phase

Everything happens somewhere

Sky

GEOAI-DataSources-SKY-01

Ground

GEOAI-DataSources-GROUND-01
GEOAI-DataSources-GROUND-UserGen-03

Below Ground

GEOAI-DataSources-BELOWGROUND-01

How to submit a GeoAI Challenge Idea?

To submit a problem statement for the GeoAI Challenge, we need the following:

1. Clear and measurable GeoAI problem statement.

2. A dataset with appropriate annotations to be shared with the participants for training and testing their AI models with suggestions on how to split the dataset for training and testing e.g. random split of 2/3rd for training 1/3rd for test, or criteria-wise splits to cover variability and diversity of data.

3. A benchmark model solution of the problem statement as an example. Videos and documentation that help new participants understand the data and problem better would also be appreciated.

4. A separate evaluation dataset to evaluate the performances of submitted solutions.

5. Cleary defined metrics for evaluation

6. Specifications for submitting solutions e.g. docker image with sample data for input and output.

7. Other relevant details about the host, timelines, etc. Please use the following MS Word template for the submission of GeoAI Challenge Statements

If you wish to discuss the potential ideas for the challenge statements and understand the requirements better, please get in touch with Lokendra Chauhan in the AIforGood GeoAI Challenge Slack Channel (coming soon).

How to participate

Benefits

geoai-ben1

Crowdsourcing multiple solutions for high-impact problems that could improve real production AI/ML systems

geoai-ben2

Increasing awareness about the problem domain and your work either in research or business

geoai-ben3

Access to a growing and highly skilled talent pool of AI researchers, students, and professionals interested in the same problems as your group for further collaborations or hiring

Sponsorship Inquiries​


    Scroll Up

    Your Gateway to #AIforGood in your inbox!

    Join our 60K+ global community to receive insights & invites to AI for Good! Just one email each week, and zero spam.