Speeding up Convolutional Neural Networks with Low Rank Expansions

Max Jaderberg, Andrea Vedaldi and Andrew Zisserman

In Proceedings British Machine Vision Conference 2014


The focus of this paper is speeding up the evaluation of convolutional neural networks. While delivering impressive results across a range of computer vision and machine learning tasks, these networks are computationally demanding, limiting their deployability. Convolutional layers generally consume the bulk of the processing time, and so in this work we present two simple schemes for drastically speeding up these layers. This is achieved by exploiting cross-channel or filter redundancy to construct a low rank basis of filters that are rank-1 in the spatial domain. Our methods are architecture agnostic, and can be easily applied to existing CPU and GPU convolutional frameworks for tuneable speedup performance. We demonstrate this with a real world network designed for scene text character recognition, showing a possible 2.5� speedup with no loss in accuracy, and 4.5� speedup with less than 1% drop in accuracy, still achieving state-of-the-art on standard benchmarks.


Poster Session


Extended Abstract (PDF, 1 page, 612K)
Paper (PDF, 13 pages, 773K)
Bibtex File


Max Jaderberg, Andrea Vedaldi, and Andrew Zisserman. Speeding up Convolutional Neural Networks with Low Rank Expansions. Proceedings of the British Machine Vision Conference. BMVA Press, September 2014.


	title = {Speeding up Convolutional Neural Networks with Low Rank Expansions},
	author = {Jaderberg, Max and Vedaldi, Andrea and Zisserman, Andrew},
	year = {2014},
	booktitle = {Proceedings of the British Machine Vision Conference},
	publisher = {BMVA Press},
	editors = {Valstar, Michel and French, Andrew and Pridmore, Tony}
	doi = { http://dx.doi.org/10.5244/C.28.88 }