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

With more and more household objects built on planned obsolescence andconsumed by a fast-growing population, hazardous waste recycling has become acritical challenge. Given the large variability of household waste, currentrecycling platforms mostly rely on human operators to analyze the scene,typically composed of many object instances piled up in bulk. Helping them byrobotizing the unitary extraction is a key challenge to speed up this tediousprocess. Whereas supervised deep learning has proven very efficient for suchobject-level scene understanding, e.g., generic object detection andsegmentation in everyday scenes, it however requires large sets of per-pixellabeled images, that are hardly available for numerous application contexts,including industrial robotics. We thus propose a step towards a practicalinteractive application for generating an object-oriented robotic grasp,requiring as inputs only one depth map of the scene and one user click on thenext object to extract. More precisely, we address in this paper the middleissue of object seg-mentation in top views of piles of bulk objects given apixel location, namely seed, provided interactively by a human operator. Wepropose a twofold framework for generating edge-driven instance segments.First, we repurpose a state-of-the-art fully convolutional object contourdetector for seed-based instance segmentation by introducing the notion ofedge-mask duality with a novel patch-free and contour-oriented loss function.Second, we train one model using only synthetic scenes, instead of manuallylabeled training data. Our experimental results show that considering edge-maskduality for training an encoder-decoder network, as we suggest, outperforms astate-of-the-art patch-based network in the present application context.

Object segmentation in depth maps with one user click and a synthetically trained fully convolutional network