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

Current learning-based robot grasping approaches exploit human-labeleddatasets for training the models. However, there are two problems with such amethodology: (a) since each object can be grasped in multiple ways, manuallylabeling grasp locations is not a trivial task; (b) human labeling is biased bysemantics. While there have been attempts to train robots using trial-and-errorexperiments, the amount of data used in such experiments remains substantiallylow and hence makes the learner prone to over-fitting. In this paper, we takethe leap of increasing the available training data to 40 times more than priorwork, leading to a dataset size of 50K data points collected over 700 hours ofrobot grasping attempts. This allows us to train a Convolutional Neural Network(CNN) for the task of predicting grasp locations without severe overfitting. Inour formulation, we recast the regression problem to an 18-way binaryclassification over image patches. We also present a multi-stage learningapproach where a CNN trained in one stage is used to collect hard negatives insubsequent stages. Our experiments clearly show the benefit of usinglarge-scale datasets (and multi-stage training) for the task of grasping. Wealso compare to several baselines and show state-of-the-art performance ongeneralization to unseen objects for grasping.

Supersizing Self-supervision: Learning to Grasp from 50K Tries and 700 Robot Hours