This is the public, feature-limited version of the conference webpage. After Registration and login please visit the full version.

Workshop: Object Representations for Learning and Reasoning

William Agnew, Rim Assouel, Michael Chang, Antonia Creswell, Eliza Kosoy, Aravind Rajeswaran, Sjoerd van Steenkiste

2020-12-11T08:00:00-08:00 - 2020-12-11T19:15:00-08:00
Abstract: Recent advances in deep reinforcement learning and robotics have enabled agents to achieve superhuman performance on a variety of challenging games and learn complex manipulation tasks. While these results are very promising, several open problems remain. In order to function in real-world environments, learned policies must be both robust to input perturbations and be able to rapidly generalize or adapt to novel situations. Moreover, to collaborate and live with humans in these environments, the goals and actions of embodied agents must be interpretable and compatible with human representations of knowledge. Hence, it is natural to consider how humans so successfully perceive, learn, and plan to build agents that are equally successful at solving real world tasks.
There is much evidence to suggest that objects are a core level of abstraction at which humans perceive and understand the world [8]. Objects have the potential to provide a compact, casual, robust, and generalizable representation of the world. Recently, there have been many advancements in scene representation, allowing scenes to be represented by their constituent objects, rather than at the level of pixels. While these works have shown promising results, there is still a lack of agreement on how to best represent objects, how to learn object representations, and how best to leverage them in agent training.
In this workshop we seek to build a consensus on what object representations should be by engaging with researchers from developmental psychology and by defining concrete tasks and capabilities that agents building on top of such abstract representations of the world should succeed at. We will discuss how object representations may be learned through invited presenters with expertise both in unsupervised and supervised object representation learning methods. Finally, we will host conversations and research on new frontiers in object learning.


To ask questions please use rocketchat, available only upon registration and login.


2020-12-11T08:00:00-08:00 - 2020-12-11T08:15:00-08:00
William Agnew
2020-12-11T08:15:00-08:00 - 2020-12-11T09:00:00-08:00
Keynote: Elizabeth Spelke
2020-12-11T09:02:00-08:00 - None
Learning Object-Centric Video Models by Contrasting Sets
2020-12-11T09:04:00-08:00 - None
Structure-Regularized Attention for Deformable Object Representation
2020-12-11T09:06:00-08:00 - None
Learning Long-term Visual Dynamics with Region Proposal Interaction Networks
2020-12-11T09:08:00-08:00 - None
Self-Supervised Attention-Aware Reinforcement Learning
2020-12-11T09:10:00-08:00 - None
Emergence of compositional abstractions in human collaborative assembly
2020-12-11T09:12:00-08:00 - None
Semantic State Representation for Reinforcement Learning
2020-12-11T09:14:00-08:00 - None
Odd-One-Out Representation Learning
2020-12-11T09:16:00-08:00 - None
Word(s) and Object(s): Grounded Language Learning In Information Retrieval
2020-12-11T09:20:00-08:00 - None
Discrete Predictive Representation for Long-horizon Planning
2020-12-11T09:22:00-08:00 - None
Dynamic Regions Graph Neural Networks for Spatio-Temporal Reasoning
2020-12-11T09:26:00-08:00 - None
Dexterous Robotic Grasping with Object-Centric Visual Affordances
2020-12-11T09:28:00-08:00 - None
Understanding designed objects by program synthesis
2020-12-11T09:29:00-08:00 - None
Learning Embeddings that Capture Spatial Semantics for Indoor Navigation
2020-12-11T09:30:00-08:00 - 2020-12-11T10:30:00-08:00
Poster Session
2020-12-11T10:30:00-08:00 - 2020-12-11T11:45:00-08:00
Panel Discussion
Jessica Hamrick
2020-12-11T11:45:00-08:00 - 2020-12-11T12:25:00-08:00
2020-12-11T12:25:00-08:00 - 2020-12-11T12:55:00-08:00
Invited Talk: Jessica Hamrick
Jessica Hamrick
2020-12-11T13:15:00-08:00 - 2020-12-11T13:25:00-08:00
Invited Talk: Irina Higgins
Irina Higgins
2020-12-11T13:25:00-08:00 - 2020-12-11T13:55:00-08:00
Invited Talk: Sungjin Ahn
Sungjin Ahn
2020-12-11T13:55:00-08:00 - 2020-12-11T14:07:00-08:00
Contributed Talk : A Symmetric and Object-Centric World Model for Stochastic Environments
2020-12-11T14:07:00-08:00 - 2020-12-11T14:19:00-08:00
Contributed Talk : OGRE: An Object-based Generalization for Reasoning Environment
2020-12-11T14:19:00-08:00 - 2020-12-11T14:49:00-08:00
Invited Talk: Wilka Carvalho
Wilka Carvalho
2020-12-11T14:49:00-08:00 - 2020-12-11T15:20:00-08:00
2020-12-11T15:20:00-08:00 - 2020-12-11T15:50:00-08:00
Invited Talk: Renée Baillargeon
2020-12-11T15:50:00-08:00 - 2020-12-12T16:20:00-08:00
Invited Talk: Dieter Fox
2020-12-11T16:20:00-08:00 - 2020-12-11T16:32:00-08:00
Contributed Talk : Disentangling 3D Prototypical Networks for Few-Shot Concept Learning
2020-12-11T16:32:00-08:00 - 2020-12-11T16:44:00-08:00
Contributed Talk : Deep Affordance Foresight: Planning for What Can Be Done Next
2020-12-11T16:44:00-08:00 - 2020-12-11T16:56:00-08:00
Contributed talk : Estimating Mass Distribution of Articulated Objects using Non-prehensile Manipulation
2020-12-11T16:56:00-08:00 - 2020-12-11T18:10:00-08:00
Klaus Greff, Josh Tenenbaum
2020-12-11T18:10:00-08:00 - 2020-12-11T18:15:00-08:00
Concluding Remarks
2020-12-11T18:15:00-08:00 - 2020-12-11T19:15:00-08:00
Poster Session B