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

Robot kinematics data, despite being a high dimensional process, is highlycorrelated, especially when considering motions grouped in certain primitives.These almost linear correlations within primitives allow us to interpret themotions as points drawn close to a union of low-dimensional linear subspaces inthe space of all motions. Motivated by results of embedding theory, inparticular, generalizations of Whitney embedding theorem, we show that randomlinear projection of motor sequences into low dimensional space loses verylittle information about structure of kinematics data. Projected points arevery good initial guess for values of latent variables in generative model forrobot sensory-motor behaviour primitives. We conducted series of experimentswhere we trained a recurrent neural network to generate sensory-motor sequencesfor robotic manipulator with 9 degrees of freedom. Experimental resultsdemonstrate substantial improvement in generalisation abilities for unobservedsamples in the case of initialization of latent variables with random linearprojection of motor data over initialization with zero or random values.Moreover, latent space is well-structured wherein samples belonging todifferent primitives are well separated from the onset of training process.

Initialization of Latent Space Coordinates via Random Linear Projections for Learning Robotic Sensory-Motor Sequences