( events)   Timezone: »  
The 2020 schedule is still incomplete Program Highlights »
Workshop
Fri Jul 17 06:00 AM -- 02:00 PM (PDT)
Workshop on Continual Learning
Haytham Fayek · Arslan Chaudhry · David Lopez-Paz · Eugene Belilovsky · Jonathan Schwarz · Marc Pickett · Rahaf Aljundi · Sayna Ebrahimi · Razvan Pascanu · Puneet Dokania





Machine learning systems are commonly applied to isolated tasks or narrow domains (e.g. control over similar robotic bodies). It is further assumed that the learning system has simultaneous access to all the data points of the tasks at hand. In contrast, Continual Learning (CL) studies the problem of learning from a stream of data from changing domains, each connected to a different learning task. The objective of CL is to quickly adapt to new situations or tasks by exploiting previously acquired knowledge, while protecting previous learning from being erased. Meeting the objectives of CL will provide an opportunity for systems to quickly learn new skills given knowledge accumulated in the past and continually extend their capabilities to changing environments, a hallmark of natural intelligence.

Introduction (Talk)
Invited Talk: Christoph H. Lampert "Learning Theory for Continual and Meta-Learning" (Talk)
Christoph H. Lampert
Live Q&A: Christoph H. Lampert (Q&A)
Spotlight Talk: Wandering Within a World: Online Contextualized Few-Shot Learning (Talk)
Invited Talk: Razvan Pascanu "Continual Learning from an Optimization/Learning-dynamics perspective" (Talk)
Razvan Pascanu
Live Q&A: Razvan Pascanu (Q&A)
Spotlight Talk: SOLA: Continual Learning with Second-Order Loss Approximation (Talk)
Invited Talk: Bing Liu "Learning on the Job in the Open World" (Talk)
Bing Liu
Live Q&A: Bing Liu (Q&A)
Spotlight Talk: Continual Learning from the Perspective of Compression (Talk)
Panel Discussion (Panel Disucssion)
Poster Session 1 (Poster session)
Invited Talk: Claudia Clopath "Continual learning though consolidation – a neuroscience angle" (Talk)
Claudia Clopath
Live Q&A: Claudia Clopath (Q&A)
Spotlight Talk: Deep Reinforcement Learning amidst Lifelong Non-Stationarity (Talk)
Invited Talk: Jeff Clune (Talk)
Jeff Clune
Live Q&A: Jeff Clune (Q&A)
Spotlight Talk: Supermasks in Superposition (Talk)
Live Invited Talk: Alexi Efros "Imagining a Post-Dataset Era" (Talk)
Alexei Efros
Live Q&A: Alexi Efros (Q&A)
Best paper: Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics (Talk)
Closing remarks (Q&A)
Poster Session 2 (Poster session)
Continual Deep Learning by Functional Regularisation of Memorable Past (Zoom Poster Session)
Online Inducing Points Selection for Gaussian Processes (Zoom Poster Session)
Task-Agnostic Continual Learning via Stochastic Synapses (Zoom Poster Session)
Routing Networks with Co-training for Continual Learning (Zoom Poster Session)
UNCLEAR: A Straightforward Method for Continual Reinforcement Learning (Zoom Poster Session)
Active Online Domain Adaptation (Zoom Poster Session)
Supermasks in Superposition (Zoom Poster Session)
Combining Variational Continual Learning with FiLM Layers (Zoom Poster Session)
Beyond Catastrophic Forgetting in Continual Learning: An Attempt with SVM (Zoom Poster Session)
Variational Beam Search for Continual Learning (Zoom Poster Session)
Online Hyperparameter Tuning for Multi-Task Learning (Zoom Poster Session)
Evaluating Agents Without Rewards (Zoom Poster Session)
Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics (Zoom Poster Session)
Variational Auto-Regressive Gaussian Processes for Continual Learning (Zoom Poster Session)
SOLA: Continual Learning with Second-Order Loss Approximation (Zoom Poster Session)
Understanding Regularisation Methods for Continual Learning (Zoom Poster Session)
Continual Reinforcement Learning with Multi-Timescale Replay (Zoom Poster Session)
Task Agnostic Continual Learning via Meta Learning (Zoom Poster Session)
Disentanglement of Color and Shape Representations for Continual Learning (Zoom Poster Session)
On Class Orderings for Incremental Learning (Zoom Poster Session)
Continual Learning from the Perspective of Compression (Zoom Poster Session)
Deep Reinforcement Learning amidst Lifelong Non-Stationarity (Zoom Poster Session)
Wandering Within a World: Online Contextualized Few-Shot Learning (Zoom Poster Session)
Understanding the Role of Training Regime in Continual Learning (Zoom Poster Session)
Visually Grounded Continual Learning of Compositional Phrases (Zoom Poster Session)
Attention Option-Critic (Zoom Poster Session)
A General Framework for Continual Learning of Compositional Structures (Zoom Poster Session)
Continual Learning in Human Activity Recognition: An empirical analysis of Regularization (Zoom Poster Session)