WALT3D: Generating Realistic Training Data from Time-Lapse Imagery for Reconstructing Dynamic Objects Under Occlusion

Current methods for 2D and 3D object understanding struggle with severe occlusions in busy urban environments, partly due to the lack of large-scale labeled ground-truth annotations for learning occlusion. In this work, we introduce a novel framework for automatically generating a large, realistic d...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Proceedings (IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Online) S. 9514 - 9524
Hauptverfasser: Vuong, Khiem, Reddy, N Dinesh, Tamburo, Robert, Narasimhan, Srinivasa G.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 16.06.2024
Schlagworte:
3D from single images
computer vision
Image segmentation
Reconstruction algorithms
Shape
Three-dimensional displays
Training
Training data
Urban areas
ISSN:1063-6919
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Current methods for 2D and 3D object understanding struggle with severe occlusions in busy urban environments, partly due to the lack of large-scale labeled ground-truth annotations for learning occlusion. In this work, we introduce a novel framework for automatically generating a large, realistic dataset of dynamic objects under occlusions using freely available time-lapse imagery. By leveraging off-the-shelf2D (bounding box, segmentation, keypoint) and 3D (pose, shape) predictions as pseudo-groundtruth, unoccluded 3D objects are identified automatically and composited into the background in a clip-art style, ensuring realistic appearances and physically accurate occlusion configurations. The resulting clip-art image with pseudogroundtruth enables efficient training of object reconstruction methods that are robust to occlusions. Our method demonstrates significant improvements in both 2D and 3D reconstruction, particularly in scenarios with heavily occluded objects like vehicles and people in urban scenes.
ISSN:1063-6919
DOI:10.1109/CVPR52733.2024.00909