LegoNet: Efficient Convolutional Neural Networks with Lego Filters

Zhaohui Yang, Yunhe Wang, Chuanjian Liu, Hanting Chen, Chunjing Xu, Boxin Shi, Chao Xu, Chang Xu
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:7005-7014, 2019.

Abstract

This paper aims to build efficient convolutional neural networks using a set of Lego filters. Many successful building blocks, e.g., inception and residual modules, have been designed to refresh state-of-the-art records of CNNs on visual recognition tasks. Beyond these high-level modules, we suggest that an ordinary filter in the neural network can be upgraded to a sophisticated module as well. Filter modules are established by assembling a shared set of Lego filters that are often of much lower dimensions. Weights in Lego filters and binary masks to stack Lego filters for these filter modules can be simultaneously optimized in an end-to-end manner as usual. Inspired by network engineering, we develop a split-transform-merge strategy for an efficient convolution by exploiting intermediate Lego feature maps. The compression and acceleration achieved by Lego Networks using the proposed Lego filters have been theoretically discussed. Experimental results on benchmark datasets and deep models demonstrate the advantages of the proposed Lego filters and their potential real-world applications on mobile devices.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-yang19c, title = {{L}ego{N}et: Efficient Convolutional Neural Networks with Lego Filters}, author = {Yang, Zhaohui and Wang, Yunhe and Liu, Chuanjian and Chen, Hanting and Xu, Chunjing and Shi, Boxin and Xu, Chao and Xu, Chang}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {7005--7014}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/yang19c/yang19c.pdf}, url = {https://proceedings.mlr.press/v97/yang19c.html}, abstract = {This paper aims to build efficient convolutional neural networks using a set of Lego filters. Many successful building blocks, e.g., inception and residual modules, have been designed to refresh state-of-the-art records of CNNs on visual recognition tasks. Beyond these high-level modules, we suggest that an ordinary filter in the neural network can be upgraded to a sophisticated module as well. Filter modules are established by assembling a shared set of Lego filters that are often of much lower dimensions. Weights in Lego filters and binary masks to stack Lego filters for these filter modules can be simultaneously optimized in an end-to-end manner as usual. Inspired by network engineering, we develop a split-transform-merge strategy for an efficient convolution by exploiting intermediate Lego feature maps. The compression and acceleration achieved by Lego Networks using the proposed Lego filters have been theoretically discussed. Experimental results on benchmark datasets and deep models demonstrate the advantages of the proposed Lego filters and their potential real-world applications on mobile devices.} }
Endnote
%0 Conference Paper %T LegoNet: Efficient Convolutional Neural Networks with Lego Filters %A Zhaohui Yang %A Yunhe Wang %A Chuanjian Liu %A Hanting Chen %A Chunjing Xu %A Boxin Shi %A Chao Xu %A Chang Xu %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-yang19c %I PMLR %P 7005--7014 %U https://proceedings.mlr.press/v97/yang19c.html %V 97 %X This paper aims to build efficient convolutional neural networks using a set of Lego filters. Many successful building blocks, e.g., inception and residual modules, have been designed to refresh state-of-the-art records of CNNs on visual recognition tasks. Beyond these high-level modules, we suggest that an ordinary filter in the neural network can be upgraded to a sophisticated module as well. Filter modules are established by assembling a shared set of Lego filters that are often of much lower dimensions. Weights in Lego filters and binary masks to stack Lego filters for these filter modules can be simultaneously optimized in an end-to-end manner as usual. Inspired by network engineering, we develop a split-transform-merge strategy for an efficient convolution by exploiting intermediate Lego feature maps. The compression and acceleration achieved by Lego Networks using the proposed Lego filters have been theoretically discussed. Experimental results on benchmark datasets and deep models demonstrate the advantages of the proposed Lego filters and their potential real-world applications on mobile devices.
APA
Yang, Z., Wang, Y., Liu, C., Chen, H., Xu, C., Shi, B., Xu, C. & Xu, C.. (2019). LegoNet: Efficient Convolutional Neural Networks with Lego Filters. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:7005-7014 Available from https://proceedings.mlr.press/v97/yang19c.html.

Related Material