DeepSym: Deep Symbol Generation and Rule Learning from Unsupervised Continuous Robot Interaction for Planning

by   Alper Ahmetoğlu, et al.

Autonomous discovery of discrete symbols and rules from continuous interaction experience is a crucial building block of robot AI, but remains a challenging problem. Solving it will overcome the limitations in scalability, flexibility, and robustness of manually-designed symbols and rules, and will constitute a substantial advance towards autonomous robots that can learn and reason at abstract levels in open-ended environments. Towards this goal, we propose a novel and general method that finds action-grounded, discrete object and effect categories and builds probabilistic rules over them that can be used in complex action planning. Our robot interacts with single and multiple objects using a given action repertoire and observes the effects created in the environment. In order to form action-grounded object, effect, and relational categories, we employ a binarized bottleneck layer of a predictive, deep encoder-decoder network that takes as input the image of the scene and the action applied, and generates the resulting object displacements in the scene (action effects) in pixel coordinates. The binary latent vector represents a learned, action-driven categorization of objects. To distill the knowledge represented by the neural network into rules useful for symbolic reasoning, we train a decision tree to reproduce its decoder function. From its branches we extract probabilistic rules and represent them in PPDDL, allowing off-the-shelf planners to operate on the robot's sensorimotor experience. Our system is verified in a physics-based 3d simulation environment where a robot arm-hand system learned symbols that can be interpreted as 'rollable', 'insertable', 'larger-than' from its push and stack actions; and generated effective plans to achieve goals such as building towers from given cubes, balls, and cups using off-the-shelf probabilistic planners.


page 4

page 5

page 6

page 8


Learning Multi-Object Symbols for Manipulation with Attentive Deep Effect Predictors

In this paper, we propose a concept learning architecture that enables a...

Probabilistic Hybrid Action Models for Predicting Concurrent Percept-driven Robot Behavior

This article develops Probabilistic Hybrid Action Models (PHAMs), a real...

Planning with Noisy Probabilistic Relational Rules

Noisy probabilistic relational rules are a promising world model represe...

Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic Boundary

Current domain-independent, classical planners require symbolic models o...

Beyond the Self: Using Grounded Affordances to Interpret and Describe Others' Actions

We propose a developmental approach that allows a robot to interpret and...

Learning High-Level Planning Symbols from Intrinsically Motivated Experience

In symbolic planning systems, the knowledge on the domain is commonly pr...

The Symbol Grounding Problem

How can the semantic interpretation of a formal symbol system be made in...