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

Grasping objects in cluttered scenarios is a challenging task in robotics.Performing pre-grasp actions such as pushing and shifting to scatter objects isa way to reduce clutter. Based on deep reinforcement learning, we propose aFast-Learning Grasping (FLG) framework, that can integrate pre-grasping actionsalong with grasping to pick up objects from cluttered scenarios with reducedreal-world training time. We associate rewards for performing moving actionswith the change of environmental clutter and utilize a hybrid triggeringmethod, leading to data-efficient learning and synergy. Then we use the outputof an extended fully convolutional network as the value function of each pixelpoint of the workspace and establish an accurate estimation of the graspprobability for each action. We also introduce a mask function as priorknowledge to enable the agents to focus on the accurate pose adjustment toimprove the effectiveness of collecting training data and, hence, to learnefficiently. We carry out pre-training of the FLG over simulated environment,and then the learnt model is transferred to the real world with minimalfine-tuning for further learning during actions. Experimental resultsdemonstrate a 94% grasp success rate and the ability to generalize to novelobjects. Compared to state-of-the-art approaches in the literature, theproposed FLG framework can achieve similar or higher grasp success rate withlesser amount of training in the real world. Supplementary video is availableat https://youtu.be/e04uDLsxfDg.

Fast-Learning Grasping and Pre-Grasping via Clutter Quantization and Q-map Masking