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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Robots learning a new manipulation task from a small amount of demonstrationsare increasingly demanded in different workspaces. A classifier model assessingthe quality of actions can predict the successful completion of a task, whichcan be used by intelligent agents for action-selection. This paper presents anovel classifier that learns to classify task completion only from a fewdemonstrations. We carry out a comprehensive comparison of different neuralclassifiers, e.g. fully connected-based, fully convolutional-based,sequence2sequence-based, and domain adaptation-based classification. We alsopresent a new dataset including five robot manipulation tasks, which ispublicly available. We compared the performances of our novel classifier andthe existing models using our dataset and the MIME dataset. The results suggestdomain adaptation and timing-based features improve success prediction. Ournovel model, i.e. fully convolutional neural network with domain adaptation andtiming features, achieves an average classification accuracy of 97.3\% and95.5\% across tasks in both datasets whereas state-of-the-art classifierswithout domain adaptation and timing-features only achieve 82.4\% and 90.3\%,respectively.

Neural Task Success Classifiers for Robotic Manipulation from Few Real Demonstrations