Faculté des Sciences appliquées
Faculté des Sciences appliquées

Master Thesis : Implicit neural representations for robotic grasping

Gustin, Julien ULiège
Promotor(s) : Louppe, Gilles ULiège
Date of defense : 26-Jun-2023/27-Jun-2023 • Permalink :
Title : Master Thesis : Implicit neural representations for robotic grasping
Translated title : [fr] Représentations neurales implicites pour la préhension robotique
Author : Gustin, Julien ULiège
Date of defense  : 26-Jun-2023/27-Jun-2023
Advisor(s) : Louppe, Gilles ULiège
Committee's member(s) : Sacré, Pierre ULiège
Bruls, Olivier ULiège
Language : English
Number of pages : 94
Keywords : [en] Deep learning
[en] Machine learning
[en] Grasping
[en] Implicit Neural Representations
[en] Prior
[en] Bayesian inference
[en] Simulation based inference
[en] Robotics
[fr] Robotique
[fr] Préhension
Discipline(s) : Engineering, computing & technology > Computer science
Commentary : A paper related to this master's thesis was published as part of a workshop entitled "Geometric Representations: The Roles of Screw Theory, Lie Algebra, & Geometric Algebra," held at ICRA 2023. Here is the link to the paper:
Target public : Researchers
Professionals of domain
General public
Institution(s) : Université de Liège, Liège, Belgique
Degree: Master en science des données, à finalité spécialisée
Faculty: Master thesis of the Faculté des Sciences appliquées


[en] Robotic grasping is a fundamental skill in many robotic applications. While most grasping methods excel in constrained tasks within structured environments. When operating in more complex and uncertain scenarios, handling uncertainty becomes essential. Bayesian frameworks provide a means to address this uncertainty but require prior knowledge about the grasping pose. However, previous research has demonstrated that using a uniform prior over the workspace is highly inefficient.

In this work, we propose a novel approach that exploits implicit neural representations to construct scene-dependent priors. This enables the application of powerful simulation-based inference algorithms to determine plausible and successful grasp poses in unstructured environments.

We demonstrate the significant improvements achieved by incorporating this informative prior. Specifically, our model achieves an impressive success rate of 97% in grasping a single object, surpassing the performance of the previous model. Additionally, we reduce acquisition time by 60% by capturing only a partial view of the scene and training an implicit neural network to reconstruct the complete scene. Furthermore, in the more complex scenario of multi-object grasping, our model achieves a success rate of 91.37% in simulation and 95.6% in real-world scenarios, comparable to benchmark models. These results demonstrate the effectiveness of our approach and its impressive sim2real transfer capabilities.

We also provide valuable explainability by examining the predicted posterior distribution. This analysis allows for a better understanding of the uncertainty associated with the estimation of the grasping pose, enhancing the transparency of the system's decision-making process.



Access master_thesis.pdf
Size: 48.64 MB
Format: Adobe PDF
Access abstract.pdf
Size: 292.03 kB
Format: Adobe PDF


  • Gustin, Julien ULiège Université de Liège > Mast. sc. don. à fin.


Committee's member(s)

  • Sacré, Pierre ULiège Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Robotique intelligente
    ORBi View his publications on ORBi
  • Bruls, Olivier ULiège Université de Liège - ULiège > Département d'aérospatiale et mécanique > Laboratoire des Systèmes Multicorps et Mécatroniques
    ORBi View his publications on ORBi
  • Total number of views 83
  • Total number of downloads 64

All documents available on MatheO are protected by copyright and subject to the usual rules for fair use.
The University of Liège does not guarantee the scientific quality of these students' works or the accuracy of all the information they contain.