ContinualAI: a Non-Profit Research Organization on Continual Learning for AI
Vincenzo Lomonaco, Keiland Cooper, Natalia Díaz Rodríguez, Timothée Lesort, German I. Parisi, Davide Maltoni, Lorenzo Pellegrini, Giacomo Bartoli, Jidin Dinesh, Manish Agnihotri, Martin Mundt
Humans have the extraordinary ability to learn continually from experience. Not only can we apply previously learned knowledge and skills to new situations, we can also use these as the foundation for later learning. One of the grand goals of Artificial Intelligence (AI) is building an artificial “Continual Learning” agent that constructs a sophisticated understanding of the world from its own experience through the autonomous incremental development of ever more complex skills and knowledge.
ContinualAI is an official non-profit research organization and the largest open community on Continual Learning for AI. We aim at connecting people and working better together on this fascinating topic we consider fundamental for the future of AI!
ContinualAI is also a striving effort to make research in continual learning and AI more open, collaborative, democratic and transparent. We support open science at every level. Everything made at ContinualAI is open and freely accessible but maybe more importantly, all the projects are open-source and collaborative since the beginning (especially in the decision processes) and reflect the needs and the ideas of the entire community, rather than an handful of elite groups.
ContinualAI: From 1 to 400+ Members!
ContinualAI was started in January 2018 by Vincenzo Lomonaco and made possible with the early contribution of Keiland Cooper and Martin Cimmino, who were finding very difficult to collaborate and discuss with other researchers working on the same topic. Information were also hard to find and gather in a single place!
Now, we are very excited to see our community grow so fast and steady, even reaching and exceeding the number of 448 members! 🔥 😮 We are always happy to add new members and get some help with our open projects or simply discuss about CL, so join us today!
Board Members and Top Contributors
ContinualAI is based on the democratic structure of a classic european non-profit association (more details here) where board members are elected every 24 months by the members of the community.
At the moment ContinualAI board members are, Vincenzo Lomonaco, Postdoctoral Researcher @ University of Bologna, Keiland Cooper, Cognitive Science & Neuroscience Researcher @ Indiana University Bloomington, Natalia Díaz Rodríguez, Assitant Professor @ ENSTA ParisTech, Timothée Lesort, PhD student @ ENSTA ParisTech and Thales, German I. Parisi, Postdoctoral Research Associate @ University of Hamburg and Davide Maltoni, Full Professor @ University of Bologna. Moreover, other 6 board members will be elected by all the members within the the end of 2019.
Exceptional members who have also greatly contributed to the ContinualAI projects include: Lorenzo Pellegrini, PhD Student @ University of Bologna, Giacomo Bartoli, Consultant @ IBM, Jidin Dinesh, CS Bachelor Student @ Cochin University of Science and Technology, Manish Agnihotri, Undergraduate Researcher @ Manipal Institute of Technology.
We are also very happy to announce an exceptional lineup in our advisory board, composed of Marc Pickett, Research Scientist @ Google AI, Eugenio Culurciello, Professor @ Purdue University, David Lopez-Paz, Research Scientist @ Facebook AI Research, Razvan Pascanu, Research Scientist @ DeepMind, Massimiliano Versace, CEO at Neurala, Subutai Ahmad, VP Research @ Numenta. All of them have shown a great expertise and interest in Continual Learning over the last few years and we are excited and honored to benefit from they precious advice!
Being grown from a single GitHub page to an official non-profit research organization, ContinualAI has now three main objectives:
- Connect: Despite the rise of the open-source adoption in the AI community, research labs work pretty much in isolation. Staying informed is also becoming more and more difficult given the speed of research in this area. Connecting researchers working on continual learning for AI is one of the reason ContinualAI exists today and one of the main goal of the organization, stimulating the exchange of information, ideas and resources on a daily basis.
- Research: In the last few years we have witnessed a renewed and steadily growing interest in the ability to learn continuously from high-dimensional data. At ContinualAI we work to create a distributed and inclusive virtual research lab on Continual Learning, where anyone can contribute and learn more about this fascinating topic, while producing cutting edge research results. While in recent years the AI community has started open-sourcing the final research products (e.g. paper and/or code), research is still conducted within closed doors in small research labs. We believe in a more inclusive approach where research is conducted openly at every stage with huge benefits for the community.
- Educate: Researchers should not only produce awesome research results but should be also concerned in making them accessible to anyone. This is why many projects are underway at ContinualAI to disseminate research in Continual Learning and related areas.
ContinualAI Open Initiatives
We are currently working together on many completely open-source/collaborative projects:
- ContinualAI VirtualLab
A collection of tools and resources that let researchers communicate and collaborate better remotely on a daily basis. Up to date, the ContinualAI Virtual Lab is based on Gitter and Slack, G Suit for Non Profit, a GitHub organization account, all the major social platforms, a Slack and a Twitter bot and much more.
- ContinualAI Wiki
A collaboratively maintained central hub of information and didactic/dissemination materials for Continual Learning and AI, hosted on Github pages and available at wiki.continualai.org
- ContinualAI Avalanche
A comprehensive framework for Continual Learning Research. It aims at unifying a set of popular CL baselines, environments and benchmarks to help algorithm prototyping and experimenting, with flexibility, reproducibility, efficiency and maintainability in mind.
- ContinualAI Colab
A number of notebooks and scripts (for demo, showcasing & tutorials) which can be directly imported into Google Co-laboratory and are related to Continual Learning.
- ContinualAI Medium Publication
At ContinualAI we value scientific dissemination. We think that to advance science it is important to promote cutting-edge research and make it accessible to a larger audience of people. This is why we maintain a Medium Publication where we try to distill and simplify the continual learning ideas with as little technical details as possible.
- ContinualAI News Mailing list and Newsletter
For people on-the-run, we also thought about a monthly newsletter with all the major breakthrough and news within this area (write us if you’d like to contribute or see your next article published!). We also maintain a Google group for the exchange of information within and outside the ContinualAI community.
- ContinualAI Papers Database
Given the amount of papers published each month on the topic within different venues, it is starting to become very difficult to keep track of the progress and the different ideas emerging in this topic. This is why one of the main projects at ContinualAI is a public papers database, with a rich-full set of attributes to help researchers find interesting papers related to their research.
- ContinualAI Short-Science
We love short-science! this is why we plan to contribute to it with a short description of as many continual learning papers as possible. Join us to be always updated about recent continual learning research progresses.
Over 500 Members and Beyond!
ContinualAI is an official non-profit research organization with a clear democratic structure with shared goals and intentions. However, we need YOUR HELP to support the ContinualAI cause! At the moment, we need new “Top Contributors” (a.k.a “Maintainers”) for the following roles (no particular experience needed, we’ll help you out! 😃):
- ContinualAI VirtualLab Maintainer.
- ContinualAI News Mailing list/Newsletter Maintainer.
- ContinualAI Colab Maintainer.
- ContinualAI Avalanche Maintainer.
- ContinualAI Website Maintainer.
- ContinualAI Wiki Maintainer.
- ContinualAI Paper Database Maintainer.
- ContinualAI Short-Science Maintainer.
- Editor and Writer for the ContinualAI Medium Publication.
ContinualAI is an official non-profit research organization and the largest open community on Continual Learning for AI. Visit our official website: continualai.org to learn more about the organization and join us as a member, or consider supporting us with a small donation! Follow us on Facebook, Twitter, Instagram, Medium, Github, YouTube!