Exportar Publicação

Vanderson Martins do Rosario, Borin, Edson & Breternitz, M. (2019). The multi-lane capsule network (MLCN). IEEE Signal Processing Letters. 26 (7), 1-1

V. M. Rosario et al.,  "The multi-lane capsule network (MLCN)", in IEEE Signal Processing Letters, vol. 26, no. 7, pp. 1-1, 2019

@article{rosario2019_1766365975368,
	author = "Vanderson Martins do Rosario and Borin, Edson and Breternitz, M.",
	title = "The multi-lane capsule network (MLCN)",
	journal = "IEEE Signal Processing Letters",
	year = "2019",
	volume = "26",
	number = "7",
	doi = "10.1109/LSP.2019.2915661",
	pages = "1-1",
	url = "https://ieeexplore.ieee.org/document/8709729"
}

TY  - JOUR
TI  - The multi-lane capsule network (MLCN)
T2  - IEEE Signal Processing Letters
VL  - 26
IS  - 7
AU  - Vanderson Martins do Rosario
AU  - Borin, Edson
AU  - Breternitz, M.
PY  - 2019
SP  - 1-1
SN  - 1070-9908
DO  - 10.1109/LSP.2019.2915661
UR  - https://ieeexplore.ieee.org/document/8709729
AB  - We introduce Multi-Lane Capsule Networks (MLCN), which are a separable and resource efficient organization of Capsule Networks (CapsNet) that allows parallel processing while achieving high accuracy at reduced cost. A MLCN is composed of a number of (distinct) parallel lanes, each contributing to a dimension of the result, trained using the routing-by-agreement organization of CapsNet. Our results indicate similar accuracy with a much-reduced cost in number of parameters for the Fashion-MNIST and Cifar10 datasets. They also indicate that the MLCN outperforms the original CapsNet when using a proposed novel configuration for the lanes. MLCN also has faster training and inference times, being more than two-fold faster than the original CapsNet in a same accelerator.
ER  -