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

The ability to interpret a scene is an important capability for a robot thatis supposed to interact with its environment. The knowledge of what is in frontof the robot is, for example, relevant for navigation, manipulation, orplanning. Semantic segmentation labels each pixel of an image with a classlabel and thus provides a detailed semantic annotation of the surroundings tothe robot. Convolutional neural networks (CNNs) are popular methods foraddressing this type of problem. The available software for training and theintegration of CNNs for real robots, however, is quite fragmented and oftendifficult to use for non-experts, despite the availability of severalhigh-quality open-source frameworks for neural network implementation andtraining. In this paper, we propose a tool called Bonnet, which addresses thisfragmentation problem by building a higher abstraction that is specific for thesemantic segmentation task. It provides a modular approach to simplify thetraining of a semantic segmentation CNN independently of the used dataset andthe intended task. Furthermore, we also address the deployment on a realrobotic platform. Thus, we do not propose a new CNN approach in this paper.Instead, we provide a stable and easy-to-use tool to make this technology moreapproachable in the context of autonomous systems. In this sense, we aim atclosing a gap between computer vision research and its use in roboticsresearch. We provide an open-source codebase for training and deployment. Thetraining interface is implemented in Python using TensorFlow and the deploymentinterface provides a C++ library that can be easily integrated in an existingrobotics codebase, a ROS node, and two standalone applications for labelprediction in images and videos.

Bonnet: An Open-Source Training and Deployment Framework for Semantic Segmentation in Robotics using CNNs