How can the #ML community contribute to #Wikipedia, solving huge challenges in vision and language? What is the importance of open and unbiased data? Do not miss the first #WikiM3L workshop tomorrow, April 29th at ICLR 2026 ! Program👉meta.wikimedia.org/wiki/Wiki-M3L/…

Get ready for the panel about "Multilinguality in multimodal research and open data" with five amazing speakers! 🎉 Moderated by Lucie-Aimée Kaffee

Learn more about the top solutions of the Wikipedia Image/Caption Matching Competition in our next session! 🏅🤩 #WikiM3L More info about the competition: diff.wikimedia.org/2021/09/13/the…

We reached the end! Thank you to all the amazing speakers, Wikipedias, PC members, and of course, the audience for all the questions! 🙏❤️ See you next year! 🎉 #WikiM3L #ICLR2022

We are excited to share our work on our Pathways Autoregressive Text-to-Image model, Parti! #Parti achieves high-fidelity photorealistic image generation and supports content-rich synthesis involving complex compositions and world knowledge. parti.research.google

Thanks a lot for the invite WikiResearch - looking forward to it and the discussions!

"Show, Interpret and Tell: Entity-aware Contextualised Image Captioning in Wikipedia" Automatically captioning Wikipedia images by integrating contextual knowledge, based on the WIT dataset. (Nguyen et al, 2022) arxiv.org/pdf/2209.10474…

Looking forward to making this TREC track come to life alongside Jheng-Hong Yang Jimmy Lin Stéphane Clinchant Miriam Redi Rafael S Rezende Krishna Srinivasan ! If you are attending TREC we will be live at 3PM ET with the introduction and preliminary discussion x.com/lintool/status…

Please check out our TREC Proposal AToMiC!

In non-ChatGPT news... AToMiC (Authoring Tools for Multimedia Content) is a new take image-text search led by Jheng-Hong Yang that will debut at TREC 2023! Check out our preprint arxiv.org/abs/2304.01961 and the track website: github.com/trec-atomic/at…

"AToMiC: An Image/Text Retrieval Test Collection to Support Multimedia Content Creation" hierarchical structures, diverse domains of texts/images, and large scale image-document associations, all from Wikipedia paper arxiv.org/abs/2304.01961 data github.com/TREC-AToMiC/AT… Jimmy Lin

A Suite of Generative Tasks for Multi-Level Multimodal Webpage Understanding abs: arxiv.org/abs/2305.03668 paper page: huggingface.co/papers/2305.03…

WikiWeb2M: A Page-Level Multimodal Wikipedia Dataset abs: arxiv.org/abs/2305.05432 paper page: huggingface.co/papers/2305.05… github: github.com/google-researc…

Interested in Cross-Modal retrieval? Feel that evaluation on caption datasets is limited and noisy? The AToMiC@TREC2023 task is here: a real, large scale problem (retrieve illustration for Wikipedia sections), with human judgements. Deadline in 45 days! Krishna Srinivasan Jimmy Lin

The final test topics will be released soon, but there’s a lot of material to start playing with" such: embeddings of openCLIP models + dev topics with human judgements over the baselines (trec-atomic.github.io/annoucements/d…).

For more information the full guidelines are available trec-atomic.github.io/annoucements/T… and organisers will be more than happy to help with any problem / answer any question (just tweet or send an email to [email protected])

We also have Jheng-Hong Yang presenting AToMiC our cross modal resource paper at #SIGIR2023 in collaboration with Jimmy Lin Miriam Redi Krishna Srinivasan Rafael S Rezende and Stéphane Clinchant

We will be presenting our work at a poster session during #SIGIR2023! Can't wait to connect with everyone in person 😎. A fantastic journey collaborating with the brilliant minds of Carlos Lassance Rafael S Rezende Krishna Srinivasan Miriam Redi Stéphane Clinchant Jimmy Lin 👉 dl.acm.org/doi/10.1145/35…

"WikiDO: A New Benchmark Evaluating Cross-Modal Retrieval for Vision-Language Models", consisting of 384K image-text pairs from Wikipedia with domain labels, along with human-verified in/out distribution test sets. (Kalyan et al, 2024) openreview.net/pdf?id=EqaSEbU…

Doing a startup and organizing taxes ... couldn't help but contemplate between AGI [1] and AGI [2]. AGI [1] = Artificial General Intelligence AGI [2] = Adjusted Gross Income