✨ Some Jumu'ah Tips
• Perform Ghusl
• Shave off your Pubic Hair and Arm Hair
• Dress nicely, and wear perfumes -not for females
• Observe 2 rak'ah of nawafil immediately you get to the Masjid
• Don't talk or greet if the sermon is on
• Recite Al-Kahf
• Observe 4 rak'ah of nawafil after the Solatul jumu'ah if in the Masjid (2 rak'ah if at home)
Some of our members were working on Indaba's pan-African communities reg; they sent me this draft for proofreading.
I blinked twice at "14 papers in Q1 2026." I thought, "No way!" But I counted them, & it was 14.
Quite humbling how productive our members are, and how modest too
“AFRICAPTION: Establishing a New Paradigm for Image Captioning,” A recent paper by Mardiyya et al., introduces a multilingual image captioning dataset covering 20 African languages, designed to move beyond English-centric benchmarks in vision-language systems.
The study which was accepted at the EACL 2026, shows that even basic tasks like image captioning remain heavily biased toward a few dominant languages and proposes a dataset that balances caption quality and length across diverse African linguistic contexts.
This is a critical development, because image captioning is a foundational task for multimodal AI systems, powering everything from accessibility tools to search, education, and assistive technologies.
But here’s the gap: Most multimodal models today are trained on datasets where language diversity is an afterthought, which means even as AI becomes more “visual,” it still struggles to speak the languages of the people it serves.
AFRICAPTION reinforces a broader shift happening in the ecosystem. We are moving from building general-purpose datasets to building context-aware, region-specific data foundations.
Because datasets like AFRICAPTION don’t just improve model performance. They redefine what representation looks like in AI systems.
This same principle guides our work as we collect speech data across African languages, even through feature phones, to power the next layer of multimodal intelligence.
Read the full paper >> https://t.co/rMvX2qrXXr
@techie_hijabie@PrinceMireku11@FatimoAdebanjo @busymindcreate @oluwatosin_olajide @J_Novikova_NLP@hemhemoh
#LanguageAI #MultimodalAI #AfricanLanguages #DatasetInfrastructure #AIInclusion #EqualyzAI
The opportunity gap in AI is more striking than ever. We talk way too much about those receiving $100M or whatever for their jobs, but not enough those asking for <$1k to present their work.
For 3rd year in a row, @ml_collective is raising funds to support @DeepIndaba attendees.
I had the privilege of learning from leading ML/AI researchers, participating in hands-on workshops on LLMs and ML math, poster session, and mentorship sessions.
Yes, the opportunity gap has less to do about talent and much to do about how said talent is developed.
@ml_collective believed. @DeepIndaba provides the platform.
Let’s help 25 people get that development.
https://t.co/AL3ZSf6rr4
@simonepoetto@lilianweng@SiyuTang3
Over a year ago, I was new to research without a clearly defined research area. It initially felt ambiguous, with little understanding of the prerequisites for conducting ML research. But @ml_collective gave me the platform to build my foundation. Although before that...
/1
We don’t just dream, we work tirelessly to solve Africa problems with the little resource we have access to. A little donation will help 25 AI researchers attend Indaba to present their works.
@AgbogaVictor@thatroblennon@AndrewYNg@james_mtc@KirkDBorne
https://t.co/7C0bD1apXG
@ml_collective has created a space for people like me and so many others to grow, ask questions, and take our first steps into ML research without feeling out of place. That kind of support means everything.
We are HALFWAY there! Thanks to all those who've kindly contributed 🙏🙏
With Indaba <4 weeks away, let's send all the 25 African researchers to their dream conference!
Donate what you can: https://t.co/ryCItIoxNs
Can curvature-aware regularization tackle the sharp minima problem in deep networks to achieve better generalization?
I hope to present my findings at DLI’25 — and we’re almost halfway there!
Help make it happen.
Donate/share: https://t.co/UHBfpWSHY2
The opportunity gap in AI is more striking than ever. We talk way too much about those receiving $100M or whatever for their jobs, but not enough those asking for <$1k to present their work.
For 3rd year in a row, @ml_collective is raising funds to support @DeepIndaba attendees.