ContinualAI Research (CLAIR) is a collaborative laboratory with the goal to foster the advance of Continual Learning (CL) technologies for AI.
Mission
Our mission is to support emerging and consolidated researchers and scientists to produce and disseminate original CL research through the organization of academic events and workshops, online meetups and reading groups, and blogpost publications.
Join our quest to investigate and answer the fascinating open-challenges of CL for AI:
- What are the key principles and mechanisms governing lifelong learning in the brain?
- How do we model artificial CL approaches that better capture the flexibility, robustness, and scalability exhibited by biological systems?
- How do we leverage current machine learning models to synergically work on embodied agents and robots that interact with the environment?
- How can we develop, advance, and deploy CL systems responsibly and ethically?
Introduction to CL
Here is a short list of papers to start with that summarize and discuss the state of the art. For a longer list of CL approaches, empirical studies, and applications, please check our wiki.
Lesort, T., Lomonaco, V., Stoian, A., Maltoni, D., Filliat, D., & Díaz-Rodríguez, N. (2020). Continual learning for robotics: Definition, framework, learning strategies, opportunities and challenges. Information Fusion Journal, 58:52-68.
Parisi, G.I., Kemker, R., Part, J.L., Kanan, C., Wermter, S. (2019) Continual Lifelong Learning with Neural Networks: A Review. Neural Networks 113:54-71 [arXiv:1802.07569]
Kemker, R., McClure, M., Abitino, A., Hayes, T.L., Kanan, C. (2018). Measuring Catastrophic Forgetting in Neural Networks. Thirty-Second AAAI Conference on Artificial Intelligence, 2018.
Díaz-Rodríguez, N., Lomonaco, V., Filliat, D., Maltoni, D. (2018) Don’t forget, there is more than forgetting: new metrics for Continual Learning. Continual Learning Workshop at NeurIPS, Montreal, Canada.
Datasets
Datasets exclusively designed for incremental/continual (robot) learning:
CORe50: A new Dataset and Benchmark for Continual Learning and Object Recognition, Detection, Segmentation.
- Lomonaco, V., Maltoni, D. (2017) CORe50: a new Dataset and Benchmark for Continuous Object Recognition. Conference on Robot Learning (CoRL), pp. 17-26.
OpenLORIS: A Dataset and Benchmark towards Lifelong Object Recognition.
- She, Q., Feng, F., Hao, X., Yang, Q., Lan, C., Lomonaco, V., Shi, X., Wang, Z., Guo, Y., Zhang, Y. and Qiao, F., (2020). OpenLORIS-Object: A Dataset and Benchmark towards Lifelong Object Recognition. International Conference on Robotics and Automation (ICRA’20).
People
CLAIR is maintained by the following ContinualAI people:
Vincenzo Lomonaco
Co-Director and Chief Scientist
Neuroscience-inspired AI, Continual Learing, Robotics
vincenzo.lomonaco@continualai.org
German I. Parisi
Co-Director and Chief Scientist
Neuroscience-inspired AI, Continual Learing, Robotics
german.parisi@continualai.org
Keiland Cooper
Research Scientist
Neuroscience, Neuroscience-inspired AI, Continual Learing
keiland.cooper@continualai.org
Qi She
Research Scientist
Continual Learing, Robotics
qi.she@continualai.org
Jeremy Forest
Research Scientist
Neuroscience, Neuroscience-inspired AI, Continual Learing
jeremy.forest@continualai.org