Skip to Main content Skip to Navigation
New interface

Détection d'objets industriels à l'aide de modèles 3D dans des images égocentriques

Julia Cohen 1 
1 imagine - Extraction de Caractéristiques et Identification
LIRIS - Laboratoire d'InfoRmatique en Image et Systèmes d'information
Abstract : Industrial manufacturing can be facilitated using innovative digital solutions such as Augmented Reality (AR). The development of new devices such as AR headsets and head-mounted devices enable operators to visualize assembly instructions while having their hands free to manipulate the physical pieces. The detection of these industrial objects through a head-mounted camera enables the virtual elements to automatically adapt to the real scene. However, images captured with an AR headset present visual artefacts inherent to the egocentric point of view. Although object detection in images is a popular application of deep learning for its effectiveness, artificial neural networks are rarely applied to egocentric images and industrial objects. The task is even more complex when no real image of the objects of interest is available, and the algorithm will be embedded in a mobile computer with a real-time inference requirement. In this thesis led in collaboration with engineering and design company DEMS, we addressed the topic of industrial objects recognition in images from an AR headset. We leveraged the available 3D models of the objects of interest in order to create a synthetic and egocentric dataset for the training of mobile and real-time neural networks. We analyzed the key elements of this synthetic dataset in order to remove the need for real images during training. Then, we proposed to use the depth information contained in RGB-D images to improve the performance of the object detector. We tackled the issue of domain generalization from synthetic to real RGB-D images, and we proposed different approaches in order to reduce the reality gap, that are compatible with a mobile and real-time inference.
Document type :
Complete list of metadata
Contributor : Julia Cohen Connect in order to contact the contributor
Submitted on : Wednesday, July 20, 2022 - 12:38:37 PM
Last modification on : Friday, September 30, 2022 - 11:34:16 AM
Long-term archiving on: : Friday, October 21, 2022 - 6:52:32 PM


Manuscrit de thèse - Julia Co...
Files produced by the author(s)


Distributed under a Creative Commons Attribution - NonCommercial - NoDerivatives 4.0 International License


  • HAL Id : tel-03728464, version 1


Julia Cohen. Détection d'objets industriels à l'aide de modèles 3D dans des images égocentriques. Intelligence artificielle [cs.AI]. LIRIS UMR 5205 CNRS/INSA de Lyon/Université Claude Bernard Lyon 1/Université Lumière Lyon 2/École Centrale de Lyon; Université Lyon 2 Lumière, 2022. Français. ⟨NNT : ⟩. ⟨tel-03728464⟩



Record views


Files downloads