In this paper, we propose a generalization of convolutional neural networks (CNN) to non-Euclidean domains for the analysis of deformable shapes. Our construction is based on localized frequency analysis (a generalization of the windowed Fourier transform to manifolds) that is used to extract the local behavior of some dense intrinsic descriptor, roughly acting as an analogy to patches in images. The resulting local frequency representations are then passed through a bank of filters whose coefficient are determined by a learning procedure minimizing a task-specific cost. Our approach generalizes several previous methods such as HKS, WKS, spectral CNN, and GPS embeddings. Experimental results show that the proposed approach allows learning class-specific shape descriptors significantly outperforming recent state-of-the-art methods on standard benchmarks.

Boscaini, D., Masci, J., Melzi, S., Bronstein, M., Castellani, U., Vandergheynst, P. (2015). Learning class-specific descriptors for deformable shapes using localized spectral convolutional networks. COMPUTER GRAPHICS FORUM, 34(5), 13-23 [10.1111/cgf.12693].

Learning class-specific descriptors for deformable shapes using localized spectral convolutional networks

MELZI, SIMONE;
2015

Abstract

In this paper, we propose a generalization of convolutional neural networks (CNN) to non-Euclidean domains for the analysis of deformable shapes. Our construction is based on localized frequency analysis (a generalization of the windowed Fourier transform to manifolds) that is used to extract the local behavior of some dense intrinsic descriptor, roughly acting as an analogy to patches in images. The resulting local frequency representations are then passed through a bank of filters whose coefficient are determined by a learning procedure minimizing a task-specific cost. Our approach generalizes several previous methods such as HKS, WKS, spectral CNN, and GPS embeddings. Experimental results show that the proposed approach allows learning class-specific shape descriptors significantly outperforming recent state-of-the-art methods on standard benchmarks.
Articolo in rivista - Articolo scientifico
Categories and Subject Descriptors (according to ACM CCS); Computational Geometry and Object Modeling [I.3.5]; Feature Measurement [I.4.7]; Learning [I.2.6];
English
13
23
11
Boscaini, D., Masci, J., Melzi, S., Bronstein, M., Castellani, U., Vandergheynst, P. (2015). Learning class-specific descriptors for deformable shapes using localized spectral convolutional networks. COMPUTER GRAPHICS FORUM, 34(5), 13-23 [10.1111/cgf.12693].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/350562
Citazioni
  • Scopus 145
  • ???jsp.display-item.citation.isi??? 123
Social impact