Please use this identifier to cite or link to this item:
https://scholarbank.nus.edu.sg/handle/10635/212722
Title: | DIFFERENTIABLE ROBOTICS: COMPOSITIONAL DEEP LEARNING WITH DIFFERENTIABLE ALGORITHM NETWORKS | Authors: | PETER KARKUS | ORCID iD: | orcid.org/0000-0002-1474-9771 | Keywords: | deep learning, planning, model-based, model-free, end-to-end, modular, neural network | Issue Date: | 1-Aug-2021 | Citation: | PETER KARKUS (2021-08-01). DIFFERENTIABLE ROBOTICS: COMPOSITIONAL DEEP LEARNING WITH DIFFERENTIABLE ALGORITHM NETWORKS. ScholarBank@NUS Repository. | Abstract: | Towards human-level robot intelligence a central question is the architecture of robot learning: how is the system represented, and how is it trained? This thesis introduces the Differentiable Algorithm Network (DAN), a compositional architecture for designing robot learning systems. The DAN is composed of neural network modules, each encoding an algorithm and associated models; and it is trained end-to-end from data. The key idea is to make model-based algorithms differentiable and encode them in a generalized neural network, thus combining the model-based modular system design with data-driven end-to-end learning. Algorithms act as structural assumptions to moderate data requirements; end-to-end learning allows modules to adapt and compensate for imperfections. We introduce DANs for a range of domains: particle filter networks for visual localization; differentiable SLAM networks for visual SLAM; QMDP networks for partially observable planning. We also combine DAN modules for visual navigation in simulation and with a real-world quadruped robot. | URI: | https://scholarbank.nus.edu.sg/handle/10635/212722 |
Appears in Collections: | Ph.D Theses (Open) |
Show full item record
Files in This Item:
File | Description | Size | Format | Access Settings | Version | |
---|---|---|---|---|---|---|
KarkusP.pdf | 7.66 MB | Adobe PDF | OPEN | None | View/Download |
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.