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

Learning state representations enables robotic planning directly from rawobservations such as images. Most methods learn state representations byutilizing losses based on the reconstruction of the raw observations from alower-dimensional latent space. The similarity between observations in thespace of images is often assumed and used as a proxy for estimating similaritybetween the underlying states of the system. However, observations commonlycontain task-irrelevant factors of variation which are nonetheless importantfor reconstruction, such as varying lighting and different camera viewpoints.In this work, we define relevant evaluation metrics and perform a thoroughstudy of different loss functions for state representation learning. We showthat models exploiting task priors, such as Siamese networks with a simplecontrastive loss, outperform reconstruction-based representations in visualtask planning.

Comparing Reconstruction- and Contrastive-based Models for Visual Task Planning