Message from President of RSJ

Welcome to Tutorial on Deep Probabilistic Generative Models for Robotics!

Objectives

The goal of this tutorial is to bring together students and researchers from robotics and machine learning to share knowledge about deep probabilistic generative models to develop future cognitive architectures for robots. In IROS 2019, we organized the workshop on deep probabilistic generative models for cognitive architecture in robotics aiming at examining the challenges and opportunities emerging from the interdisciplinary research field covering machine learning, cognitive science, and robotics. In the workshop, we investigated how we could create a cognitive architecture for a robot using deep and probabilistic generative models. We shared knowledge about the state-of-the-art machine learning methods that contribute to model language-related capabilities in robotics, and to exchange views among cutting-edge robotics researchers with a special emphasis on the usage of deep generative models in robotics and modeling a wide range of cognitive capabilities using probabilistic generative models.

Based on this experience, a tutorial on the usage of “pixyz” as well as the basic theories of deep generative models will be provided in order to advance the research utilizing the discussions last year. Pixyz is a high-level deep generative modeling library, based on PyTorch. It is developed with a focus on enabling easy implementation of various deep generative models (https://github.com/masasu/pixyz). Knowing the usage of this tool will be a big advantage for implementing the deep generative models.

We will also give some comprehensive examples of applications of probabilistic generative models for robotics followed by the tutorial on “Serket.” Serket is another useful tools for integrating different types of probabilistic generative models, which is useful for developing a large scale cognitive model for robots (http://serket.naka-lab.org/).

We believe the topic of this tutorial is timely and necessary for the IEEE-RAS community.

The targeted audience includes:

• graduate students and robotics researchers who are interested in machine learning and deep learning (especially probabilistic generative models),

• graduate students and robotics researchers who are interested in building integrative cognitive architecture using deep generative models

• graduate students and robotics researchers who are interested in the integration of high-level and low-level cognitive capabilities in a robot,

• graduate students and robotics researchers who are interested in hierarchical Bayesian modeling of cognitive capabilities,

• graduate students, and machine learning and artificial intelligence researchers who are interested in the application of deep probabilistic generative models and multimodal machine learning to robotics,

• people in industry that are trying to develop real-world application using robotic and machine learning technologies.

Since this year's IROS will be held in an "on-demand" manner, this tutorial will be included in the "on-demand" system. To join this tutorial, please register on IROS 2020 On-Demand.

For those who already registered on IROS 2020, you have already been automatically signed-up for On-Demand; you have nothing else to do. All others must sign-up:

Visit IROS On-Demand Sign-Up to register for free access

Registration for IROS 2020 On-Demand is free!

IROS On-Demand (goes live Oct 25, 2020) is the platform to view pre-recorded videos of the 1400+ Technical Talks, Plenaries and Keynotes, and 60+ Workshops and Tutorials. The contents of this tutorial will be featured on On-Demand as pre-recorded content. All tutorials will be free to view. Again, we emphasize that this will be pre-recorded content; there are no Slack, Zoom or other synchronous channels. On-Demand reinforces the approach of accessing content anytime, anywhere and with any device.

Organizers

• Takayuki Nagai Osaka University

• Tadahiro Taniguchi Ritsumeikan University

• Takato Horii Osaka University

• Chie Hieida Nara Institute of Science and Technology

• Kaede Hayashi Ritsumeikan University

Endorsements

• IEEE RAS TC on Robot Learning

• IEEE RAS TC on Cognitive Robotics

• IEEE CDS TC Task Force on Robotics