English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this work we augment a Deep Q-Learning agent with a Reward Machine (DQRM)to increase speed of learning vision-based policies for robot tasks, andovercome some of the limitations of DQN that prevent it from converging togood-quality policies. A reward machine (RM) is a finite state machine thatdecomposes a task into a discrete planning graph and equips the agent with areward function to guide it toward task completion. The reward machine can beused for both reward shaping, and informing the policy what abstract state itis currently at. An abstract state is a high level simplification of thecurrent state, defined in terms of task relevant features. These twosupervisory signals of reward shaping and knowledge of current abstract statecoming from the reward machine complement each other and can both be used toimprove policy performance as demonstrated on several vision based robotic pickand place tasks. Particularly for vision based robotics applications, it isoften easier to build a reward machine than to try and get a policy to learnthe task without this structure.

Disentangled Planning and Control in Vision Based Robotics via Reward Machines