Show simple item record

dc.rights.licensehttps://creativecommons.org/licenses/by-sa/2.5/ar/es_AR
dc.contributor.authorFainstein, Migueles_AR
dc.contributor.authorSiless, Vivianaes_AR
dc.contributor.authorIarussi, Emmanueles_AR
dc.date.accessioned2024-02-19T14:43:35Z
dc.date.available2024-02-19T14:43:35Z
dc.date.issued2024
dc.identifier.urihttps://repositorio.utdt.edu/handle/20.500.13098/12380
dc.description.abstractIn recent years, there has been a growing interest in training Neural Networks to approximate Unsigned Distance Fields (UDFs) for representing open surfaces in the context of 3D reconstruction. However, UDFs are nondifferentiable at the zero level set which leads to significant errors in distances and gradients, generally resulting in fragmented and discontinuous surfaces. In this paper, we propose to learn a hyperbolic scaling of the unsigned distance field, which defines a new Eikonal problem with distinct boundary conditions. This allows our formulation to integrate seamlessly with state-of-the-art continuously differentiable implicit neural representation networks, largely applied in the literature to represent signed distance fields. Our approach not only addresses the challenge of open surface representation but also demonstrates significant improvement in reconstruction quality and training performance. Moreover, the unlocked field’s differentiability allows the accurate computation of essential topological properties such as normal directions and curvatures, pervasive in downstream tasks such as rendering. Through extensive experiments, we validate our approach across various data sets and against competitive baselines. The results demonstrate enhanced accuracy and up to an order of magnitude increase in speed compared to previous methods.es_AR
dc.format.extent14 p.es_AR
dc.format.mediumapplication/pdfes_AR
dc.languageenges_AR
dc.publisherUniversidad Torcuato Di Tellaes_AR
dc.rightsinfo:eu-repo/semantics/openAccesses_AR
dc.subjectNeural Networkses_AR
dc.subjectRedes Neuronaleses_AR
dc.subject3D reconstructiones_AR
dc.subjectHyperbolic scalinges_AR
dc.subjectOpen surface representationes_AR
dc.subjectRedes de representación neural implícitaes_AR
dc.subjectImplicit neural representation networkses_AR
dc.titleDUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scalinges_AR
dc.typeinfo:eu-repo/semantics/articlees_AR
dc.subject.keywordUnsigned Distance Fields (UDFs)es_AR
dc.subject.keywordEikonal problemes_AR
dc.subject.keywordBoundary conditionses_AR
dc.subject.keywordCampos de distancia no firmados (UDFs)es_AR
dc.subject.keywordEscalado hiperbólicoes_AR
dc.subject.keywordProblema eikonales_AR
dc.type.versioninfo:eu-repo/semantics/publishedVersiones_AR


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record