SS3DNet-AF: A Single-Stage, Single-View 3D Reconstruction Network with Attention-Based Fusion

Learning object shapes from a single image is challenging due to variations in scene content, geometric structures, and environmental factors, which create significant disparities between 2D image features and their corresponding 3D representations, hindering the effective training of deep learning...

Ausführliche Beschreibung

Bibliografische Detailangaben

Hauptverfasser:	Muhammad Awais Shoukat, Allah Bux Sargano, Alexander Malyshev, Lihua You, Zulfiqar Habib
Format:	Artigo
Sprache:	Inglês
Veröffentlicht:	MDPI AG 2024-12-01
Schriftenreihe:	Applied Sciences
Schlagworte:	SS3DNet-AF 3D reconstruction attention-based fusion point clouds
Online-Zugang:	https://www.mdpi.com/2076-3417/14/23/11424
Tags:	Tag hinzufügen Keine Tags, Fügen Sie das erste Tag hinzu!