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

Assistive robots should be able to wash, fold or iron clothes. However, dueto the variety, deformability and self-occlusions of clothes, creatinggeneral-purpose robot systems for cloth manipulation is challenging. Syntheticdata is a promising direction to improve generalization, though its usabilityis often limited by the sim-to-real gap. To advance the use of synthetic datafor cloth manipulation and to enable tasks such as robotic folding, we presenta synthetic data pipeline to train keypoint detectors for almost flattenedcloth items. To test its performance, we have also collected a real-worlddataset. We train detectors for both T-shirts, towels and shorts and obtain anaverage precision of 64.3%. Fine-tuning on real-world data improves performanceto 74.2%. Additional insight is provided by discussing various failure modes ofthe keypoint detectors and by comparing different approaches to obtain clothmeshes and materials. We also quantify the remaining sim-to-real gap and arguethat further improvements to the fidelity of cloth assets will be required tofurther reduce this gap. The code, dataset and trained models are availableonline.

Learning Keypoints for Robotic Cloth Manipulation using Synthetic Data