東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Algorithmic Optimization of First Co...

Chamarthi, Ramachandra Vikas.

FindBook

Google Book

Amazon

博客來

Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design./
作者:	Chamarthi, Ramachandra Vikas.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2019,
面頁冊數:	66 p.
附註:	Source: Masters Abstracts International, Volume: 80-10.
Contained By:	Masters Abstracts International80-10.
標題:	Computer Engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13806248
ISBN:	9781392029664

Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design.
Chamarthi, Ramachandra Vikas.

Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design. - Ann Arbor : ProQuest Dissertations & Theses, 2019 - 66 p.

Source: Masters Abstracts International, Volume: 80-10.

Thesis (M.S.)--The University of North Carolina at Charlotte, 2019.

This item must not be sold to any third party vendors.

This thesis proposes "1D Convolution replacement layer", a novel optimization for first convolution layer in CNN. This optimization enables edge friendly streaming accelerator design with a minimum drop in accuracy. This optimization reduces the number of convolution parameters in the first convolution layers of CNN, reducing the number of multiplications performed in convolution operation. In CNNs first convolution is the most memory and compute intensive as the first layer operates on input. In a streaming accelerator design, the first layer operates on streaming data, the complexity of operations and memory demand of the first layer will proportionally affect the latency of complete accelerator design. Using 1D Convolution replacement in a CNN on a N x N convolution layer after 1D replacement number of operations in each convolution window gets reduced by N times. To show the effect of 1D convolution in accelerators, streaming accelerator design for SqueezeNet is compared with 1D-SqueezeNet, SqueezeNet with 1D convolution replacement in the first layer is discussed. 1D replacement enabled edge friendly design reducing the dynamic power consumption by 7.3X, with 0.6\\% drop in accuracy in SqueezeNet real-time edge accelerator.

ISBN: 9781392029664Subjects--Topical Terms:

1567821
Computer Engineering.

Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design.
LDR:02324nmm a2200325 4500 001 2208153
005 20190929184225.5
008 201008s2019 ||||||||||||||||| ||eng d
020 $a 9781392029664
035 $a (MiAaPQ)AAI13806248
035 $a (MiAaPQ)uncc:11979
035 $a AAI13806248
040 $a MiAaPQ $c MiAaPQ
100 1 $a Chamarthi, Ramachandra Vikas. $3 3435166
245 1 0 $a Algorithmic Optimization of First Convolution Layer in Cnns for Hardware Accelerator Design.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2019
300 $a 66 p.
500 $a Source: Masters Abstracts International, Volume: 80-10.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Tabkhi, Hamed.
502 $a Thesis (M.S.)--The University of North Carolina at Charlotte, 2019.
506 $a This item must not be sold to any third party vendors.
520 $a This thesis proposes "1D Convolution replacement layer", a novel optimization for first convolution layer in CNN. This optimization enables edge friendly streaming accelerator design with a minimum drop in accuracy. This optimization reduces the number of convolution parameters in the first convolution layers of CNN, reducing the number of multiplications performed in convolution operation. In CNNs first convolution is the most memory and compute intensive as the first layer operates on input. In a streaming accelerator design, the first layer operates on streaming data, the complexity of operations and memory demand of the first layer will proportionally affect the latency of complete accelerator design. Using 1D Convolution replacement in a CNN on a N x N convolution layer after 1D replacement number of operations in each convolution window gets reduced by N times. To show the effect of 1D convolution in accelerators, streaming accelerator design for SqueezeNet is compared with 1D-SqueezeNet, SqueezeNet with 1D convolution replacement in the first layer is discussed. 1D replacement enabled edge friendly design reducing the dynamic power consumption by 7.3X, with 0.6\\% drop in accuracy in SqueezeNet real-time edge accelerator.
590 $a School code: 0694.
650 4 $a Computer Engineering. $3 1567821
650 4 $a Electrical engineering. $3 649834
690 $a 0464
690 $a 0544
710 2 $a The University of North Carolina at Charlotte. $b Electrical Engineering. $3 3170016
773 0 $t Masters Abstracts International $g 80-10.
790 $a 0694
791 $a M.S.
792 $a 2019
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13806248