Human-centric partitioning of the environment

Karaoğuz, Hakan; Bore, Nils; Folkesson, John

doi:10.1109/roman.2017.8172401

Cited by 2 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the last years deep neural networks [6] have become the new dominant learning paradigm in visual recognition, establishing the new state of the art in various visual tasks such as object classification [7] and object detection [8]. Similarly deep architectures have been applied on real robots [9], [10], leading to significant improvements on a variety of robot vision tasks [11], [12], [13]. One known challenge with DNNs is that they are data hungry.…”

Section: Online Adaptationmentioning

confidence: 99%

Kitting in the Wild through Online Domain Adaptation

Mancini

Karaoğuz

Ricci

et al. 2018

2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Self Cite

View full text Add to dashboard Cite

Technological developments call for increasing perception and action capabilities of robots. Among other skills, vision systems that can adapt to any possible change in the working conditions are needed. Since these conditions are unpredictable, we need benchmarks which allow to assess the generalization and robustness capabilities of our visual recognition algorithms. In this work we focus on robotic kitting in unconstrained scenarios. As a first contribution, we present a new visual dataset for the kitting task. Differently from standard object recognition datasets, we provide images of the same objects acquired under various conditions where camera, illumination and background are changed. This novel dataset allows for testing the robustness of robot visual recognition algorithms to a series of different domain shifts both in isolation and unified. Our second contribution is a novel online adaptation algorithm for deep models, based on batchnormalization layers, which allows to continuously adapt a model to the current working conditions. Differently from standard domain adaptation algorithms, it does not require any image from the target domain at training time. We benchmark the performance of the algorithm on the proposed dataset, showing its capability to fill the gap between the performances of a standard architecture and its counterpart adapted offline to the given target domain.

show abstract