A Transformer‐Based Network for Full Object Pose Estimation with Depth Refinement

zu Verbundenen Objekten

In response to increasing demand for robotics manipulation, accurate vision‐based full pose estimation is essential. While convolutional neural networks‐based approaches have been introduced, the quest for higher performance continues, especially for precise robotics manipulation, including in the Agri‐robotics domain. This article proposes an improved transformer‐based pipeline for full pose estimation, incorporating a Depth Refinement Module. Operating solely on monocular images, the architecture features an innovative Lighter Depth Estimation Network using a Feature Pyramid with an up‐sampling method for depth prediction. A Transformer‐based Detection Network with additional prediction heads is employed to directly regress object centers and predict the full poses of the target objects. A novel Depth Refinement Module is then utilized alongside the predicted centers, full poses, and depth patches to refine the accuracy of the estimated poses. The performance of this pipeline is extensively compared with other state‐of‐the‐art methods, and the results are analyzed for fruit picking applications. The results demonstrate that the pipeline improves the accuracy of pose estimation to up to 90.79% compared to other methods available in the literature.

Standort
Deutsche Nationalbibliothek Frankfurt am Main
Umfang
Online-Ressource
Sprache
Englisch

Erschienen in
A Transformer‐Based Network for Full Object Pose Estimation with Depth Refinement ; day:28 ; month:07 ; year:2024 ; extent:14
Advanced intelligent systems ; (28.07.2024) (gesamt 14)

Urheber
Abdulsalam, Mahmoud
Ahiska, Kenan
Aouf, Nabil

DOI
10.1002/aisy.202400110
URN
urn:nbn:de:101:1-2407291408567.364705458377
Rechteinformation
Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.
Letzte Aktualisierung
14.08.2025, 10:52 MESZ

Datenpartner

Dieses Objekt wird bereitgestellt von:
Deutsche Nationalbibliothek. Bei Fragen zum Objekt wenden Sie sich bitte an den Datenpartner.
Original beim Datenpartner anzeigen

Beteiligte

  • Abdulsalam, Mahmoud
  • Ahiska, Kenan
  • Aouf, Nabil

Ähnliche Objekte (12)

Improving Object Grasp Performance via Transformer-Based Sparse Shape Completion

Improving Object Grasp Performance via Transformer-Based Sparse Shape Completion

Transformer-Based Context Condensation for Boosting Feature Pyramids in Object Detection

Transformer-Based Context Condensation for Boosting Feature Pyramids in Object Detection

Object detection using convolutional neural networks and transformer-based models: a review

Object detection using convolutional neural networks and transformer-based models: a review

Transformer-based convolutional forgetting knowledge tracking

Transformer-based convolutional forgetting knowledge tracking

A transformer-based Urdu image caption generation

A transformer-based Urdu image caption generation

Temporal fusion transformer-based prediction in aquaponics

Temporal fusion transformer-based prediction in aquaponics

Transformer-based tool recommendation system in Galaxy

Transformer-based tool recommendation system in Galaxy

Transformer-based NMT : modeling, training and implementation

Transformer-based NMT : modeling, training and implementation

BertSRC: transformer-based semantic relation classification

BertSRC: transformer-based semantic relation classification

Transformer-based models for multimodal irony detection

Transformer-based models for multimodal irony detection

Image2SMILES: Transformer‐Based Molecular Optical Recognition Engine **

Image2SMILES: Transformer‐Based Molecular Optical Recognition Engine **

Transformer-based tool recommendation system in Galaxy

Transformer-based tool recommendation system in Galaxy

Verbundene Objekte