CORDEIL, Maxime; BILLY, Thomas; MELLADO, Nicolas; BARTHE, Loic; COUTURE, Nadine; REUTER, Patrick

doi:10.1109/ISMAR-Adjunct60411.2023.00025

Voir/Ouvrir

ESTIA_ISMAR_2023_Cordeil.pdf (3.601Mo)

Métadonnées

Licence d’utilisation du document

CORDEIL, Maxime
The University of Queensland [UQ [All campuses : Brisbane, Dutton Park Gatton, Herston, St Lucia and other locations]]

BILLY, Thomas
The University of Queensland [UQ [All campuses : Brisbane, Dutton Park Gatton, Herston, St Lucia and other locations]]

MELLADO, Nicolas
Centre National de la Recherche Scientifique [CNRS]
Structural Models and Tools in Computer Graphics [IRIT-STORM]

Langue

Communication dans un congrès

Ce document a été publié dans

ISMAR 2023 - IEEE International Symposium on Mixed and Augmented Reality Adjunct, IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct 2023), 2023-10-16, Sydney. p. 81-85

IEEE

Résumé en anglais

We propose an initial exploration of an interactive machine-learning (IML) dialogue in immersive, interactive Virtual Reality (VR) for the classification of points in 3D point clouds. We contribute ImmersiveIML, an Immersive Analytics tool which builds on humanmachine learning trial-and-error dialogue to support an iterative classification process of points in the 3D point cloud. The interactions in ImmersiveIML are designed to be both expressive and minimal; we designed the iterative process to be supported by (1) 6-DOF controllers to allow direct interaction in large 3D point clouds via a minimal brushing technique, (2) a fast trainable machine learning model, and (3) the direct visual feedback of classification results. This constitutes an iterative human-in-the-loop process that eventually converges to a classification according to human intent. We argue that this approach is a novel contribution that supports a constant improvement of the classification model and fast tracks classification tasks with this type of data, in an immersive scenario. We report on the design and implementation of ImmersiveIML and demonstrate its capabilities with two emblematic application scenarios: edge detection in 3D and classification of trees in a city LiDAR dataset.< Réduire

Mots clés en anglais

Gestural input

Interaction paradigms

Virtual reality

Computing methodologies

Neural networks

Human-centered computing

Visualization

Visu- alization techniques

Interaction techniques

URI

https://oskar-bordeaux.fr/handle/20.500.12278/187687

DOI

http://dx.doi.org/10.1109/ISMAR-Adjunct60411.2023.00025

Project ANR

Espace d'échelles pour l'apprentissage machine sur des nuages de points 3d - ANR-22-CE23-0004

Unités de recherche

ESTIA - Recherche