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

Classical Visual Servoing (VS) rely on handcrafted visual features, whichlimit their generalizability. Recently, a number of approaches, some based onDeep Neural Networks, have been proposed to overcome this limitation bycomparing directly the entire target and current camera images. However, bygetting rid of the visual features altogether, those approaches require thetarget and current images to be essentially similar, which precludes thegeneralization to unknown, cluttered, scenes. Here we propose to perform VSbased on visual features as in classical VS approaches but, contrary to thelatter, we leverage recent breakthroughs in Deep Learning to automaticallyextract and match the visual features. By doing so, our approach enjoys theadvantages from both worlds: (i) because our approach is based on visualfeatures, it is able to steer the robot towards the object of interest even inpresence of significant distraction in the background; (ii) because thefeatures are automatically extracted and matched, our approach can easily andautomatically generalize to unseen objects and scenes. In addition, we proposeto use a render engine to synthesize the target image, which offers a furtherlevel of generalization. We demonstrate these advantages in a robotic graspingtask, where the robot is able to steer, with high accuracy, towards the objectto grasp, based simply on an image of the object rendered from the camera viewcorresponding to the desired robot grasping pose.

DFBVS: Deep Feature-Based Visual Servo