Title | ||
---|---|---|
Designing by Training - Acceleration Neural Network for Fast High-Dimensional Convolution. |
Abstract | ||
---|---|---|
The high-dimensional convolution is widely used in various disciplines but has a serious performance problem due to its high computational complexity. Over the decades, people took a handmade approach to design fast algorithms for the Gaussian convolution. Recently, requirements for various non-Gaussian convolutions have emerged and are continuously getting higher. However, the handmade acceleration approach is no longer feasible for so many different convolutions since it is a time-consuming and painstaking job. Instead, we propose an Acceleration Network (AccNet) which turns the work of designing new fast algorithms to training the AccNet. This is done by: 1, interpreting splatting, blurring, slicing operations as convolutions; 2, turning these convolutions to gCP layers to build AccNet. After training, the activation function g together with AccNet weights automatically define the new splatting, blurring and slicing operations. Experiments demonstrate AccNet is able to design acceleration algorithms for a ton of convolutions including Gaussian/non-Gaussian convolutions and produce state-of-the-art results. |
Year | Venue | Keywords |
---|---|---|
2018 | ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 31 (NIPS 2018) | fast algorithms,neural network,activation function |
Field | DocType | Volume |
Mathematical optimization,Activation function,Convolution,Computer science,Slicing,Algorithm,Gaussian,Acceleration,Artificial neural network,Computational complexity theory | Conference | 31 |
ISSN | Citations | PageRank |
1049-5258 | 0 | 0.34 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Longquan Dai | 1 | 23 | 2.73 |
Liang Tang | 2 | 45 | 14.11 |
Yuan Xie | 3 | 407 | 27.48 |
Jinhui Tang | 4 | 5180 | 212.18 |