Meet the winners of UmojaHack #1: SAEON Marine Invertebrates Identification Challenge

Deep sea marine invertebrates are frequently long-lived, slow growing and have limited to zero mobility. As a result of these life history characteristics they are known to reflect the changing climate of the ecosystem. They frequently have wide tolerance ranges as they are unable to rapidly move away from unfavourable conditions. For this reason, marine invertebrates are considered good indicators of long-term ecosystem health. In a changing ocean environment it is likely that marine invertebrates will provide early warning signs through reduced biodiversity, along with dramatic increases of a few species for which the changes are beneficial.

Marine invertebrates living on South Africa’s seafloor are meticulously collected, measured and photographed annually during research trawl surveys. Individual invertebrates are photographed in isolation, especially if identification thereof is uncertain, enabling identification at a later stage. Identification of each individual in each image is a laborious effort and is still required. Once the species have been identified, the data are added to a South African marine invertebrates database. Any species not previously encountered are also being added to a field guide to improve taxonomic knowledge of the region.

Teams are challenged to develop an automated image classification solution for photographs of marine invertebrate taken by researchers in South Africa. This will substantially reduce the manual image processing efforts of the team and enable them to detect any changing patterns in marine invertebrates much faster by reducing the need for human intervention in sample processing and evaluation.

Watch a SAEON team out in the field collecting data.

About South African Environmental Observation Network (SAEON) (http://www.saeon.ac.za/):

SAEON is a sustained, coordinated, responsive and comprehensive in situ South African Earth observation network that delivers long-term reliable data for scientific research and informs decision-making for a knowledge society and improved quality of life.

Thank you to Microsoft and African Bank for sponsoring UmojaHack.

You may use only the datasets provided for this competition. Your solution must use machine learning.

This is a computer vision challenge. No metadata may be used.

If your solution places 1st, 2nd, or 3rd in the final ranking, you will be required to submit your winning solution code to us for verification and you thereby agree to share all worldwide rights of copyright in and to such winning solution to Zindi.

If your solution places 1st, 2nd, or 3rd in the final ranking on the private leaderboard, you will be required to submit your winning solution code to us for verification.

You will have until 18:29 GMT on 21 March 2020 to submit your code for review. Submit your code to zindi@zindi.africa with subject line “Challenge name position # - team name or username and university.” Regardless of any public announcement of winners, Zindi reserves the right to disqualify any user, team, or university on or even after 21 March 2020 if the code does not reproduce the winning submission.

Individual competitors are able to form teams on this competition. You must accept the rules associated with teaming up.

Multiple accounts per user are not allowed. Collaboration across individuals not in the same team is not allowed. And collaboration across different teams that is not allowed.

Code must not be shared privately. Any code that is shared, must be made available to all competition participants through the platform.

The solution must use publicly-available, open-source packages only.

Maximum 200 solutions submitted per day. Your highest-scoring solution will be the one by which you are judged.

Note that there are Public and Private Leaderboards. The Public Leaderboard excludes approximately 50% of the test dataset. While the competition is open, the Public Leaderboard will rank the submitted solutions by the accuracy score they achieve. Upon close of the competition, the Private Leaderboard, which covers 100% of the test dataset, will be made public and will constitute the final ranking for the competition.

The data used in this competition is the sole property of Zindi and the competition host. You may not transmit, duplicate, publish, redistribute or otherwise provide or make available any competition data to any party not participating in the Competition (this includes uploading the data to any public site such as Kaggle or GitHub). You may upload, store and work with the data on any cloud platform such as Google Colab, AWS or similar, as long as 1) the data remains private and 2) doing so does not contravene Zindi’s rules of use.

If two solutions earn identical scores on the leaderboard, the tie breaker will be the date and time in which the submission was made (the earlier solution will win).

Refer to the FAQs and Terms of Use for additional rules that may apply to this competition.

You acknowledge and agree that Zindi may, without any obligation to do so, remove or disqualify an individual, team, account, or university if Zindi believes that such individual, team, account, or university is in violation of these Rules.

Teams may win in one challenge category, and are encouraged to enter only one

• Participants may compete individually or in teams of up to four people.
• The teams will be judged based on their ranking on the dedicated Zindi leaderboard at the time of competition close.
• All participants in the hackathon must be registered students (undergraduate or graduate) at the university they represent. Lecturers, University staff, and alumni may participate in a mentorship or advisory capacity.
• Teams cannot collaborate or share information with each other.
• All solutions must use machine learning, but teams are permitted and encouraged to use exploratory data analysis in building their solutions.
• All solutions must use publicly-available, open-source packages only.
• Solutions must use only the allowed and available datasets.
• Participants caught cheating or breaking any competition rules will be immediately disqualified from the competition.
• Universities caught cheating or allowing teams to cheat will be immediately disqualified from the competition.
• The winning code must be submitted to Zindi for review and validation immediately at the close of the competition. In the interest of logistics, code review will take place only after the competition has closed and winners have been announced.

We reserve the right to modify these rules at any time as necessary.

The evaluation metric for this challenge is Log Loss.

Note that there are 137 classes (species of invertebrate). Values should be probabilities and can be between 0 and 1 inclusive. The probabilities do not need to add up to 1.

Your submission file should look like:

FILE           Pteraster_capensis   Porifera   Astropecten_irregularis_pontoporeus ...
00G9CO1.jpeg            1                0                    0
01TO3K4.jpeg            0.98             0.33                 0.78                                         
01YAQRV.jpeg            0.10             0.56                 0.21      

In order to win, you must:

• be a part of the UmojaHack Africa event on 21 March 2020
• be currently enrolled as a student at an African university
• have your affiliated university listed on your Zindi profile

1st prize: $1,000 USD shared between members of the winning team plus LinkedIn Learning access and Microsoft certification vouchers (can be used for the AZ-900: Microsoft Azure Fundamentals or Azure Data Scientist Associate Certificate) for all team members

2nd prize: LinkedIn Learning access and Microsoft certification vouchers(can be used for the AZ-900: Microsoft Azure Fundamentals or Azure Data Scientist Associate Certificate) for all team members

3rd prize: LinkedIn Learning access and Microsoft certification vouchers (can be used for the AZ-900: Microsoft Azure Fundamentals or Azure Data Scientist Associate Certificate) for all team members

All winners’ names and photos will appear on Zindi social media accounts.

08:00 – 08:30 GMT Welcome and orientation (live video conference across all university locations using Zoom and YouTube)

08:30 – 09:00 GMT Technical orientation to the platform and the challenges (live video conference across all university locations using Zoom and YouTube)

09:00 GMT Competition opens (note that users can sign up for a competition and join teams before the time)

12:00 - 13:00 GMT: Spot prizes and meet SAEON and Xente

09:00 – 16:29 GMT Students form teams and work on the challenge, questions and issues during this time can be addressed by local Zindi reps or via WhatsApp group

16:29 GMT Submissions close

16:45 – 17:15 GMT Announcement of international winners and prizes (video conference across all university locations via Zoom and YouTube)

17:15 - 17:30 GMT Announcement of local winners and prizes

Welcome to all the students from:

Abubakar Tafawa Balewa University

Adama Science & Technology University

African Institute for Mathematical Sciences (AIMS-Senegal)

African Leadership University

Ashesi University

Bayero University Kano

Benson Idahosa University

Bindura University of Zimbabwe

Blossom Academy

Cairo University

Carnegie Mellon University Africa (CMU-Africa)

Central University

Chinhoyi University of Technology

Dakar Institute of Technology

Daystar University

Dedan Kimathi University of Technology

Durban University of Technology

University of Sousse (ENISo)

Enugu State University of Science and Technology

ESI - Ecole Supérieure d'Informatique Alger

Federal University of Agriculture, Abeokuta

Federal University of Petroleum Resources, Effurun

Federal University of Technology Akure, FUTA

Federal University of Technology Minna

Federal University of Technology, Owerri

Federal University Oye Ekiti

Ghana Technology University College

Higher School Of Communication of Tunis (SUPCOM)

Information and Communications University

IT Business School

Jomo Kenyatta University of Agriculture and Technology

KCA University

Kenyatta University

Kwame Nkrumah University of Science and Technology

Kwame Nkrumah University of Science and Technology

Ladoke Akintola University of Technology (LAUTECH)

Ladoke Akintola University of Technology Ogbomoso

Lagos State University

Maasai Mara University

Makerere University

Meru University of Science and Technology

Michael Okpara University of Agriculture

Moringa School

MultimediaUniversity of Kenya

National Advanced School of Engineering (NASE)

National Engineering School of Tunis (ENIT)

National School of Electronics and Telecommunications of Sfax

National University of Science and Technology

Nelson Mandela African Institution of Science and Technology (NM-AIST)

Nnamdi Azikiwe University

Obafemi Awolowo University Ile-ife

Olabisi Onabanjo University

Stellenbosch University

The Technical University of Kenya (TU-K)

The University Of Dodoma

Tshwane University of Technology

Uganda Technology and Management University

United States International University-Africa

University for Development Studies

University of Benin

University of Cape Coast

University of Cape Town

University of Dar es Salaam

University of Energy and Natural Resources

University of Ghana

University of Ibadan

University of Ilorin

University of Lagos

University of Limpopo

University of Malawi The Polytechnic

University of Sol Plaatje

University of Uyo

Vaal University of Technology