Invertible Neural Warp for NeRF

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Chng, Shin-Fang, Garg, Ravi, Saratchandran, Hemanth, Lucey, Simon
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929424322002944
author Chng, Shin-Fang
Garg, Ravi
Saratchandran, Hemanth
Lucey, Simon
author_facet Chng, Shin-Fang
Garg, Ravi
Saratchandran, Hemanth
Lucey, Simon
contents This paper tackles the simultaneous optimization of pose and Neural Radiance Fields (NeRF). Departing from the conventional practice of using explicit global representations for camera pose, we propose a novel overparameterized representation that models camera poses as learnable rigid warp functions. We establish that modeling the rigid warps must be tightly coupled with constraints and regularization imposed. Specifically, we highlight the critical importance of enforcing invertibility when learning rigid warp functions via neural network and propose the use of an Invertible Neural Network (INN) coupled with a geometry-informed constraint for this purpose. We present results on synthetic and real-world datasets, and demonstrate that our approach outperforms existing baselines in terms of pose estimation and high-fidelity reconstruction due to enhanced optimization convergence.
format Preprint
id arxiv_https___arxiv_org_abs_2407_12354
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Invertible Neural Warp for NeRF
Chng, Shin-Fang
Garg, Ravi
Saratchandran, Hemanth
Lucey, Simon
Computer Vision and Pattern Recognition
This paper tackles the simultaneous optimization of pose and Neural Radiance Fields (NeRF). Departing from the conventional practice of using explicit global representations for camera pose, we propose a novel overparameterized representation that models camera poses as learnable rigid warp functions. We establish that modeling the rigid warps must be tightly coupled with constraints and regularization imposed. Specifically, we highlight the critical importance of enforcing invertibility when learning rigid warp functions via neural network and propose the use of an Invertible Neural Network (INN) coupled with a geometry-informed constraint for this purpose. We present results on synthetic and real-world datasets, and demonstrate that our approach outperforms existing baselines in terms of pose estimation and high-fidelity reconstruction due to enhanced optimization convergence.
title Invertible Neural Warp for NeRF
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.12354