DWM: A Decomposable Winograd Method for Convolution Acceleration

نویسندگان
چکیده

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Improving accuracy of Winograd convolution for DNNs

Modern deep neural networks (DNNs) spend a large amount of their execution time computing convolutions. Winograd's minimal algorithm for small convolutions can greatly reduce the number of arithmetic operations. However, a large reduction in floating point (FP) operations in these algorithms can result in significantly reduced FP accuracy of the result. In this paper we propose several methods ...

متن کامل

Enabling Sparse Winograd Convolution by Native Pruning

Sparse methods and the use of Winograd convolutions are two orthogonal approaches, each of which significantly accelerates convolution computations in modern CNNs. Sparse Winograd merges these two and thus has the potential to offer a combined performance benefit. Nevertheless, training convolution layers so that the resulting Winograd kernels are sparse has not hitherto been very successful. B...

متن کامل

Pruning of Winograd and FFT Based Convolution Algorithm

Winogradand FFT-based convolution are two efficient convolution algorithms targeting high-performance inference. Their efficiency comes from the reduction of the number of multiplication operations due to linear and Fourier transforms. However, the two existing approaches cannot handle efficient compression of the neural network, which might contribute significant improvement in computation and...

متن کامل

On the Coppersmith–Winograd method

Until a few years ago, the fastest known matrix multiplication algorithm, due to Coppersmith and Winograd (1990), ran in time O(n). Recently, a surge of activity by Stothers, Vassilevska-Williams and Le Gall has led to an improved algorithm running in time O(n), due to Le Gall (2014). These algorithms are obtained by analyzing higher and higher tensor powers of a certain identity of Coppersmith...

متن کامل

A Discrete Singular Convolution Method for the Seepage Analysis in Porous Media with Irregular Geometry

A novel discrete singular convolution (DSC)  formulation  is  presented for the seepage analysis in irregular geometric porous media. The DSC is a new promising numerical approach which has been recently applied to solve several engineering problems. For a medium with regular geometry, realizing of the DSC for the seepage analysis is straight forward. But DSC implementation for a medium with ir...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Proceedings of the AAAI Conference on Artificial Intelligence

سال: 2020

ISSN: 2374-3468,2159-5399

DOI: 10.1609/aaai.v34i04.5838