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

Approaching robotic cloth manipulation using reinforcement learning based onvisual feedback is appealing as robot perception and control can be learnedsimultaneously. However, major challenges result due to the intricate dynamicsof cloth and the high dimensionality of the corresponding states, what shadowsthe practicality of the idea. To tackle these issues, we propose TraKDis, anovel Transformer-based Knowledge Distillation approach that decomposes thevisual reinforcement learning problem into two distinct stages. In the firststage, a privileged agent is trained, which possesses complete knowledge of thecloth state information. This privileged agent acts as a teacher, providingvaluable guidance and training signals for subsequent stages. The second stageinvolves a knowledge distillation procedure, where the knowledge acquired bythe privileged agent is transferred to a vision-based agent by leveragingpre-trained state estimation and weight initialization. TraKDis demonstratesbetter performance when compared to state-of-the-art RL techniques, showing ahigher performance of 21.9%, 13.8%, and 8.3% in cloth folding tasks insimulation. Furthermore, to validate robustness, we evaluate the agent in anoisy environment; the results indicate its ability to handle and adapt toenvironmental uncertainties effectively. Real robot experiments are alsoconducted to showcase the efficiency of our method in real-world scenarios.

TraKDis: A Transformer-based Knowledge Distillation Approach for Visual Reinforcement Learning with Application to Cloth Manipulation