CADTrack: Instructions and Support for Orientation Disambiguation of Near-Symmetrical Objects

TY - ABST

T1 - CADTrack

T2 - ACM Interactive Surfaces and Spaces

AU - Evangelista Belo, Joao Marcelo

AU - Wissing, Jon

AU - Feuchtner, Tiare

AU - Grønbæk, Kaj

PY - 2023/11

Y1 - 2023/11

N2 - Determining the correct orientation of objects can be critical to succeed in tasks like assembly and quality assurance. In particular, near-symmetrical objects may require careful inspection of small visual features to disambiguate their orientation. We propose CADTrack, a digital assistant for providing instructions and support for tasks where the orientation of near-symmetrical objects matters. Additionally, we present a deep learning pipeline for tracking the orientation of near-symmetrical objects. In contrast to existing approaches which require labeled datasets involving laborious data acquisition and annotation processes, CADTrack uses a digital model of the object to generate synthetic data and train a convolutional neural network. Furthermore, we extend the architecture of Mask R-CNN with a confidence prediction branch to avoid errors caused by misleading orientation guidance. We evaluate CADTrack through a user study, comparing our tracking-based instructions to other methods, confirming the benefits of our approach in terms of preference and lower

AB - Determining the correct orientation of objects can be critical to succeed in tasks like assembly and quality assurance. In particular, near-symmetrical objects may require careful inspection of small visual features to disambiguate their orientation. We propose CADTrack, a digital assistant for providing instructions and support for tasks where the orientation of near-symmetrical objects matters. Additionally, we present a deep learning pipeline for tracking the orientation of near-symmetrical objects. In contrast to existing approaches which require labeled datasets involving laborious data acquisition and annotation processes, CADTrack uses a digital model of the object to generate synthetic data and train a convolutional neural network. Furthermore, we extend the architecture of Mask R-CNN with a confidence prediction branch to avoid errors caused by misleading orientation guidance. We evaluate CADTrack through a user study, comparing our tracking-based instructions to other methods, confirming the benefits of our approach in terms of preference and lower

KW - dataset generation

KW - deep learning

KW - guidance

KW - industry 4.0

KW - near-symmetrical objects

KW - tracking

KW - user study

UR - http://www.scopus.com/inward/record.url?scp=85177558459&partnerID=8YFLogxK

U2 - 10.1145/3626462

DO - 10.1145/3626462

M3 - Conference abstract in proceedings

T3 - Proceedings of the ACM on Human-Computer Interaction

BT - Proceedings of the ACM on Human-Computer Interaction

A2 - Mäkelä, Ville

A2 - Lucero, Andrés

A2 - Alt, Florian

A2 - Hancock, Mark

PB - Association for Computing Machinery

CY - New York

Y2 - 5 November 2023 through 8 November 2023

ER -