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-based approaches to robotic manipulation are limited by thescalability of data collection and accessibility of labels. In this paper, wepresent a multi-task domain adaptation framework for instance grasping incluttered scenes by utilizing simulated robot experiments. Our neural networktakes monocular RGB images and the instance segmentation mask of a specifiedtarget object as inputs, and predicts the probability of successfully graspingthe specified object for each candidate motor command. The proposed transferlearning framework trains a model for instance grasping in simulation and usesa domain-adversarial loss to transfer the trained model to real robots usingindiscriminate grasping data, which is available both in simulation and thereal world. We evaluate our model in real-world robot experiments, comparing itwith alternative model architectures as well as an indiscriminate graspingbaseline.

Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from Simulation