12 Oct 2022

Mintirho ya Vulavula: Deep Learning Indaba #4 - An Indaba deferred, but one that will shape our collective destiny

Reflecting on the 2022 Deep Learning Indaba held in Tunis, Tunisia - Prof Vukosi Marivate

Mintirho ya Vulavula - Good work, speaks for itself - Actions speak louder than words

Let us fast forward from 2019 https://dsfsi.github.io/blog/go-tsamaya-ke-go-bona/ to 2022.

A pandemic, a dream deferred and a changed world

What happened in between? COVID-19 happened and the African ML community innovated their way through it. 2020 onwards has seen great strides in the African ML community. We had wider participation of Africans in many top flight conferences due to the move to online. This removed some of the challenges of travelling or costs of registration to these conferences. As we move back to physical conferences, we should reflect on how we can improve participation or reach of our top venues.

Many grassroots organisations launched new initiatives.

AI4D worked through 2020 to provide resources for the creation of NLP datasets.
Deep Learning Indaba and Data Science Africa launched joint research grants in 2021;
Masakhane led the hosting of 3 Africa NLP workshops;
Deep Learning Indaba began its mentorship program;
BlackInAI pushed;
And many many more.

Still we yearned for that return to the physical meeting of the Indaba. In 2022, that opportunity would come back. The dream deferred, but not forgotten. Tunis was the homecoming destination after almost 3 years of no Deep Learning Indaba.

Tunis, here we come

Tunisia was the venue for the 2022 Deep Learning Indaba. We are all thankful for the Higher School of Communication of Tunis (Ecole supérieure des communications de Tunis) for being the host venue for 400 or more members of the Deep Learning Indaba Community. For some context, the Indaba typically takes place over a 6 day period and this year was similar.

A high level view of the schedule:

Day 0 - Sunday: Arrivals, Registration, Introductory Tutorials, Welcome Event
Day 1 & 2 - Monday & Tuesday: Plenary, Invited Keynotes, Practicals
Day 3 - Wednesday: Africa Research Showcase, Community Lightning Talks, Poster Sessions
Day 4 - Thursday: Parallel Workshops, Invited Keynotes
Day 5 - Thursday: Parallel Workshops, Panels, Closing Plenary, Closing Event

The schedule is designed in such a way that there are a lot of opportunities for people to engage and also choose their own path. From beginners to seasoned researchers, one can find something for their interest.

Day 3 is always a highlight as many of the participants present posters of their research and the halls are a hive of activity as one experiences the poster session.

Highlights for DSFSI

DSFSI has always supported the indaba through organising, tutoring and participation. Below are some of the contributions to the 2022 program.

POSTERS

Masters students, Matimba and Thapelo presented posters at this year’s Indaba. The titles of the posters are

Xitsonga <>English translation using NMT with Back translation Poster Link
Cross-Lingual Embedding Methods and Applications: A Systematic Review for Low-Resourced Scenarios Poster Link

You can learn more about our research via our publications here

Vukosi was one of the organisers who coordinated the poster judging for best posters for Indaba 2022.

PRACTICALS

T he Indaba is known for high quality practical sessions. Abiodun facilitated and tutored at the practical sessions.

PRESENTATIONS AND PANELS

Vukosi represented the Lacuna fund at the NLP workshop on Thursday. Vukosi was a panellist on the data.org Panel Discussion on Friday evening.

Masakhane celebrated its 3rd Birthday!!!

The @MasakhaneNLP birthday party at the NLP Workshop at the @DeepIndaba. 🥳🎉
It's amazing to see how much the community has grown over the last 3 years! pic.twitter.com/eqmGART9PK
— Sebastian Ruder (@seb_ruder) August 25, 2022

Experiences from DSFSI Members

Matimba Shingange -

“If we are talking about life-changing experiences, Deep Learning Indaba 2022 was for me one of those experiences. My intention to attend the next conference was clear from the very first day of the conference. The event is well organised and offers an excellent opportunity for people from different walks of life to connect and learn. Every day offered me the opportunity to learn about what other people do and share what I do. Every break or walk to the next venue was an opportunity to meet and talk to someone new. The opportunity to also get to tour Tunisia in between was also amazing. Having the opportunity to attend Deep Learning Indaba was an amazing experience and one I am very grateful for.”

So, where to next?

Thank you, thank you, thank you.

First, we want to thank the local organising committee for such a great event and we know they likely did not sleep for a long while before and a short while after. We are also thankful to the expanded Deep Learning Indaba organising committee and leadership. The strive to innovate forward is very much appreciated and apparent in the program and interactions at the indaba.

Think X, IndabaX

In the 2022 edition of the Indaba, the research day was coordinated by IndabaX leaders. The IndabaX program was created to spur on smaller localised communities across the African continent through them hosting local IndabaX events. Officially - “A Deep Learning IndabaX is a locally-organised Indaba that helps spread knowledge and builds capacity in machine learning.” The IndabaX program, held in over 20 countries. This means the IndabaX program reaches more participants than the main Indaba. From June to December 2022, there will be IndabaX events. Check for one near you https://deeplearningindaba.com/2022/indabax/

Get Involved

As an individual

Do great work. The Indaba aims to strengthen African Machine Learning. This is through YOU being a critical contributor, owner, and shaper of the coming advances in artificial intelligence and machine learning. As such we need you to do great work and share it with the world.

Build the community and others. Connect with your local and continental community. There are many organisations like DLI, Data Science Africa, Data Science Network, Masakhane, Sisonke Biotik, GhanaNLP, etc. Work to join and contribute

Spread the word. Let others know that African Machine Learning is here. Talk about the work of others. Refer people.

As a company or a government entity

Support the Deep Learning Indaba. The Indaba takes many volunteers and resources to put together. We need to have strong funding partners to keep the initiative going. Visit https://deeplearningindaba.com/2022/indaba/sponsorship/ to see how you can get involved and also interact with the community at each Indaba.

Support or grow AI Research and Development on the continent. There is a need for organisations across the continent to increase of R&D so that ultimately the African continent’s total R&D spend per capita increases. The DLI welcomes discussions on sponsorship and also ideas on other ways for organisations to get involved.