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

Transformers have revolutionized vision and natural language processing withtheir ability to scale with large datasets. But in robotic manipulation, datais both limited and expensive. Can manipulation still benefit from Transformerswith the right problem formulation? We investigate this question with PerAct, alanguage-conditioned behavior-cloning agent for multi-task 6-DoF manipulation.PerAct encodes language goals and RGB-D voxel observations with a PerceiverTransformer, and outputs discretized actions by ``detecting the next best voxelaction''. Unlike frameworks that operate on 2D images, the voxelized 3Dobservation and action space provides a strong structural prior for efficientlylearning 6-DoF actions. With this formulation, we train a single multi-taskTransformer for 18 RLBench tasks (with 249 variations) and 7 real-world tasks(with 18 variations) from just a few demonstrations per task. Our results showthat PerAct significantly outperforms unstructured image-to-action agents and3D ConvNet baselines for a wide range of tabletop tasks.

Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation