Stable Low-Rank Tensor Decomposition for Compression of Convolutional Neural Network

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985556%3A_____%2F20%3A00534541" target="_blank" >RIV/67985556:_____/20:00534541 - isvavai.cz</a>

Result on the web

<a href="http://dx.doi.org/10.1007/978-3-030-58526-6_31" target="_blank" >http://dx.doi.org/10.1007/978-3-030-58526-6_31</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-030-58526-6_31" target="_blank" >10.1007/978-3-030-58526-6_31</a>

Result language

angličtina

Original language name

Stable Low-Rank Tensor Decomposition for Compression of Convolutional Neural Network

Original language description

Most state-of-the-art deep neural networks are overparameterized and exhibit a high computational cost. A straightforward approach to this problem is to replace convolutional kernels with its low-rank tensor approximations, whereas the Canonical Polyadic tensor Decomposition is one of the most suited models. However, fitting the convolutional tensors by numerical optimization algorithms often encounters diverging components, i.e.,extremely large rank-one tensors but canceling each other. Such degeneracy often causes the non-interpretable result and numerical instability for the neural network ne-tuning. This paper is the first study on degeneracy in the tensor decomposition of convolutional kernels. We present a novel method, which can stabilize the low-rank approximation of convolutional kernels and ensure efficient compression while preserving the high quality performance of the neural networks. We evaluate our approach on popular CNN architectures for image classification and show that our method results in much lower accuracy degradation and provides consistent performance.n

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

20201 - Electrical and electronic engineering

Project

—

Continuities

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2020

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

ECCV 2020

ISBN

978-3-030-58525-9

ISSN

0302-9743

e-ISSN

1611-3349

Number of pages

18

Pages from-to

522-539

Publisher name

Springer Nature Switzerland AG 2020

Place of publication

Cham

Event location

Glasgow

Event date

Aug 23, 2020

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—