Learning Implicit Functions for Dense 3D Shape Correspondence of Generic Objects

Liu, Feng; Liu, Xiaoming

doi:10.1109/tpami.2022.3233431

Cited by 2 publications

(1 citation statement)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Non-rigid 3D registration, correspondence matching and 3DMM fitting are highly-active research areas. Methods 979-8-3503-4544-5/23/$31.00 ©2023 IEEE focus on organic shapes such as faces [14], [8], [19], human bodies [16], [10], [20] and various human organs [12] or man made objects, such as chairs, cups and aircraft [24], [15]. A current popular approach is to employ implicit surface representations; for example, where the 3D surface is the zero level set of a learnt Signed Distance Function [18], [9].…”

Section: Related Workmentioning

confidence: 99%

Laplacian ICP for Progressive Registration of 3D Human Head Meshes

Pears

Dai

Smith

et al. 2023

2023 IEEE 17th International Conference on Automatic Face and Gesture Recognition (FG)

View full text Add to dashboard Cite

We present a progressive 3D registration framework that is a highly-efficient variant of classical non-rigid Iterative Closest Points (N-ICP). Since it uses the Laplace-Beltrami operator for deformation regularisation, we view the overall process as Laplacian ICP (L-ICP). This exploits a 'small deformation per iteration' assumption and is progressively coarse-to-fine, employing an increasingly flexible deformation model, an increasing number of correspondence sets, and increasingly sophisticated correspondence estimation. Correspondence matching is only permitted within predefined vertex subsets derived from domain-specific feature extractors. Additionally, we present a new benchmark and a pair of evaluation metrics for 3D non-rigid registration, based on annotation transfer. We use this to evaluate our framework on a publicly-available dataset of 3D human head scans (Headspace). The method is robust and only requires a small fraction of the computation time compared to the most popular classical approach, yet has comparable registration performance.

show abstract