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

We present a data-efficient framework for solving visuomotor sequentialdecision-making problems which exploits the combination of reinforcementlearning (RL) and latent variable generative models. Our framework trains deepvisuomotor policies by introducing an action latent variable such that thefeed-forward policy search can be divided into three parts: (i) training asub-policy that outputs a distribution over the action latent variable given astate of the system, (ii) unsupervised training of a generative model thatoutputs a sequence of motor actions conditioned on the latent action variable,and (iii) supervised training of the deep visuomotor policy in an end-to-endfashion. Our approach enables safe exploration and alleviates thedata-inefficiency problem as it exploits prior knowledge about valid sequencesof motor actions. Moreover, we provide a set of measures for evaluation ofgenerative models such that we are able to predict the performance of the RLpolicy training prior to the actual training on a physical robot. We define twonovel measures of disentanglement and local linearity for assessing the qualityof latent representations, and complement them with existing measures forassessment of the learned distribution. We experimentally determine thecharacteristics of different generative models that have the most influence onperformance of the final policy training on a robotic picking task.

Data-efficient visuomotor policy training using reinforcement learning and generative models