We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: GroSS: Group-Size Series Decomposition for Grouped Architecture Search

Abstract: We present a novel approach which is able to explore the configuration of grouped convolutions within neural networks. Group-size Series (GroSS) decomposition is a mathematical formulation of tensor factorisation into a series of approximations of increasing rank terms. GroSS allows for dynamic and differentiable selection of factorisation rank, which is analogous to a grouped convolution. Therefore, to the best of our knowledge, GroSS is the first method to enable simultaneous training of differing numbers of groups within a single layer, as well as all possible combinations between layers. In doing so, GroSS is able to train an entire grouped convolution architecture search-space concurrently. We demonstrate this through architecture searches with performance objectives on multiple datasets and networks. GroSS enables more effective and efficient search for grouped convolutional architectures.
Comments: Accepted for publication at ECCV 2020
Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)
Cite as: arXiv:1912.00673 [cs.LG]
  (or arXiv:1912.00673v3 [cs.LG] for this version)

Submission history

From: Henry Howard-Jenkins [view email]
[v1] Mon, 2 Dec 2019 10:32:50 GMT (119kb,D)
[v2] Mon, 23 Mar 2020 12:26:25 GMT (4399kb,D)
[v3] Thu, 16 Jul 2020 16:28:12 GMT (4659kb,D)

Link back to: arXiv, form interface, contact.