Making Sense of Vision and Touch: Learning Multimodal Representations for Contact-Rich Tasks

by   Michelle A. Lee, et al.

Contact-rich manipulation tasks in unstructured environments often require both haptic and visual feedback. It is non-trivial to manually design a robot controller that combines these modalities which have very different characteristics. While deep reinforcement learning has shown success in learning control policies for high-dimensional inputs, these algorithms are generally intractable to deploy on real robots due to sample complexity. In this work, we use self-supervision to learn a compact and multimodal representation of our sensory inputs, which can then be used to improve the sample efficiency of our policy learning. Evaluating our method on a peg insertion task, we show that it generalizes over varying geometries, configurations, and clearances, while being robust to external perturbations. We also systematically study different self-supervised learning objectives and representation learning architectures. Results are presented in simulation and on a physical robot.


page 1

page 8

page 13


Making Sense of Vision and Touch: Self-Supervised Learning of Multimodal Representations for Contact-Rich Tasks

Contact-rich manipulation tasks in unstructured environments often requi...

Proactive Action Visual Residual Reinforcement Learning for Contact-Rich Tasks Using a Torque-Controlled Robot

Contact-rich manipulation tasks are commonly found in modern manufacturi...

From Visual Place Recognition to Navigation: Learning Sample-Efficient Control Policies across Diverse Real World Environments

Visual navigation tasks in real world environments often require both se...

Deep Reinforcement Learning for Contact-Rich Skills Using Compliant Movement Primitives

In recent years, industrial robots have been installed in various indust...

SEMI: Self-supervised Exploration via Multisensory Incongruity

Efficient exploration is a long-standing problem in reinforcement learni...

Learning Dense Rewards for Contact-Rich Manipulation Tasks

Rewards play a crucial role in reinforcement learning. To arrive at the ...

Impact Makes a Sound and Sound Makes an Impact: Sound Guides Representations and Explorations

Sound is one of the most informative and abundant modalities in the real...