Few-camera Dynamic Scene Variational Novel-view Synthesis - Archive ouverte HAL Access content directly
Conference Papers Year : 2020

Few-camera Dynamic Scene Variational Novel-view Synthesis

Beatrix-Emőke Fülöp-Balogh
Nicolas Bonneel
Julie Digne

Abstract

Few-camera videos Structure from motion Novel depth and RGB Relaxed point reconstruction without temporal consistency Efficient pose estimation robust to dynamic objects Sparse noisy point clouds Camera poses Virtual camera path Variational optimization with temporal consistency Figure 1: Given a set of video sequences from a few cameras only, our method computes camera poses and sparse points, then optimizes those points into a novel video sequence following a user-defined camera path. Our space-time SfM relaxes temporal consistency for points on dynamic objects and instead robustly adds consistency during the novel view synthesis via our variational formulation.
Fichier principal
Vignette du fichier
JFIG2020___Consistent_Depth_Estimation_for_Dynamic_Scene_Multi_view_Synthesis (1).pdf (22.81 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03372486 , version 1 (10-10-2021)

Identifiers

  • HAL Id : hal-03372486 , version 1

Cite

Beatrix-Emőke Fülöp-Balogh, Eleanor Tursman, Nicolas Bonneel, James Tompkin, Julie Digne. Few-camera Dynamic Scene Variational Novel-view Synthesis. JOURNÉES FRANÇAISES D'INFORMATIQUE GRAPHIQUE, Nov 2020, Nancy, France. ⟨hal-03372486⟩
69 View
15 Download

Share

Gmail Facebook X LinkedIn More