2021-08-16 10:00:44 索煒達(dá)電子 1149
項(xiàng)目編號(hào):E164
文件大?。?0M
源碼說(shuō)明:帶中文注釋
開(kāi)發(fā)環(huán)境:C編譯器
簡(jiǎn)要概述:
關(guān)于我們的本科畢業(yè)項(xiàng)目,使用 Vivado High Level Synthesis 2016.4 和 Vivado SDSoC 2016.4 實(shí)現(xiàn) LeNet-5
Win 10 測(cè)試應(yīng)用
您可以通過(guò)自己的手寫(xiě)數(shù)字圖像測(cè)試加速器
如果您想測(cè)試該應(yīng)用程序,請(qǐng)按照以下說(shuō)明操作
Configure the IP address of Zedboard.
username@Zedboard:~# ifconfig
Start .elf file with port name argument (in here, 5555 is port name)
username@Zedboard:~# lenet5_test.elf 5555
Start the win 10 test application and input the IP address & port name.
Press connect
Open image file
I did not put a zoom in/out function to the app, so please suit the image size.
Used model is LeNet5-Like Deep CNN
Input : -1.0 to 1.0
Conv1 : 1x32x32 -> 6x28x28, ksize = 1x6x5x5, stride = 1
Pool1 : 6x28x28 -> 6x14x14, average pooling, window size = 2x2, stride = 2
Conv2 : 6x14x14 -> 16x10x10, ksize = 6x16x25, stride = 1
Pool2 : 16x10x10 -> 16x5x5, average pooling, window size = 2x2, stride = 2
Conv3 : 16x5x5 -> 120x1x1, ksize = 16x120x25, stride = 1
FC1 : 120x84
FC2 : 84x10
I used Zedboard(Zynq 7z020) for testing.
HW Functions : CONVOLUTION_ LAYER_ 1, CONVOLUTION_ LAYER_ 2, and CONVOLUTION_ LAYER_ 3, Clk freq set as 100MHz.
SW accuracy : 98.63% (single precision fp) HW accuracy : 98.63% (single precision fp)
# of images : 10,000, batch size : 1 SW runtime : 59.4456 seconds HW runtime : 16.3954 seconds speedup : x3.63 faster
Changwoo Lee (Hanyang University, Seoul, South Korea)
Jeonghyun Woo (Hanyang University, Seoul, South Korea)
文件列表:
目錄│文件列表:
└ lenet5_hls
└ lenet5_hls
│ LOG.h
│ main.cpp
│ sdx_test.h
├ filter
│ │ bconv1.mdl
│ │ bconv3.mdl
│ │ bconv5.mdl
│ │ bfc1.mdl
│ │ bfc2.mdl
│ │ bpool1.mdl
│ │ bpool2.mdl
│ │ Wconv1.mdl
│ │ Wconv3.mdl
│ │ Wconv3_modify.mdl
│ │ Wconv5.mdl
│ │ Wfc1.mdl
│ │ Wfc2.mdl
│ │ Wpool1.mdl
│ │ Wpool2.mdl
│ └ new
│ │ bconv1.mdl
│ │ bconv3.mdl
│ │ bconv5.mdl
│ │ bfc1.mdl
│ │ bfc2.mdl
│ │ bpool1.mdl
│ │ bpool2.mdl
│ │ LeNet-weights_Conv_1.txt
│ │ LeNet-weights_Conv_1_Bias.txt
│ │ LeNet-weights_Conv_2.txt
│ │ LeNet-weights_Conv_2_Bias.txt
│ │ LeNet-weights_Conv_2_Dummy.txt
│ │ LeNet-weights_Conv_2_Dummy.txt.bak
│ │ LeNet-weights_Conv_3.txt
│ │ LeNet-weights_Conv_3_Bias.txt
│ │ LeNet-weights_Fc_1.txt
│ │ LeNet-weights_Fc_1_Bias.txt
│ │ LeNet-weights_Fc_2.txt
│ │ LeNet-weights_Fc_2.txt.bak
│ │ LeNet-weights_Fc_2_Bias.txt
│ │ LeNet-weights_Fc_2_Bias.txt.bak
│ │ LeNet-weights_Pool_1.txt
│ │ LeNet-weights_Pool_1_Bias.txt
│ │ LeNet-weights_Pool_2.txt
│ │ LeNet-weights_Pool_2_Bias.txt
│ │ Wconv1.mdl
│ │ Wconv3.mdl
│ │ Wconv5.mdl
│ │ Wfc1.mdl
│ │ Wfc2.mdl
│ │ Wpool1.mdl
│ └ Wpool2.mdl
├ lenet5
│ │ classify_lib.h
│ │ common.h
│ │ lenet5.h
│ ├ hw_layers
│ │ │ activation.cpp
│ │ │ activation.h
│ │ │ image_convolution.cpp
│ │ │ image_convolution.h
│ │ │ image_fullyconnected.h
│ │ │ image_pool.cpp
│ │ └ image_pool.h
│ └ sw_layers
│ │ image_convolution_sw.h
│ │ image_fullyconnected_sw.h
│ └ image_pool_sw.h
├ MNIST_DATA
│ │ MNIST_DATA.h
│ │ t10k-images.idx3-ubyte
│ └ t10k-labels.idx1-ubyte
├ Win10 Test App
│ └ LeNet5 Test
│ │ LeNet5 Test.sln
│ ├ .vs
│ │ └ LeNet5 Test
│ │ └ v15
│ │ │ .suo
│ │ ├ Server
│ │ │ └ sqlite3
│ │ │ │ db.lock
│ │ │ └ storage.ide
│ │ └ sqlite3
│ │ │ db.lock
│ │ └ storage.ide
│ └ LeNet5 Test
│ │ App.xaml
│ │ App.xaml.cs
│ │ LeNet5 Test.csproj
│ │ LeNet5 Test.csproj.user
│ │ LeNet5 Test_TemporaryKey.pfx
│ │ MainPage.xaml
│ │ MainPage.xaml.cs
│ │ Package.appxmanifest
│ ├ Assets
│ │ │ a-2729794_960_720.png
│ │ │ LockScreenLogo.scale-200.png
│ │ │ picture-frame-with-mountain-image_318-40293.jpg
│ │ │ SplashScreen.scale-200.png
│ │ │ Square150x150Logo.scale-200.png
│ │ │ Square44x44Logo.scale-200.png
│ │ │ Square44x44Logo.targetsize-24_altform-unplated.png
│ │ │ star.png
│ │ │ StoreLogo.png
│ │ └ Wide310x150Logo.scale-200.png
│ ├ BundleArtifacts
│ │ │ arm.txt
│ │ │ x64.txt
│ │ └ x86.txt
│ ├ obj
│ │ │ LeNet5 Test.csproj.nuget.cache
│ │ │ LeNet5 Test.csproj.nuget.g.props
│ │ │ LeNet5 Test.csproj.nuget.g.targets
│ │ └ project.assets.json
│ └ Properties
│ │ AssemblyInfo.cs
│ └ Default.rd.xml
└ Win10 Test App Release
│ LeNet5 Test_1.0.0.0_Test.zip
└ README.md