Deep Learning toolbox commands sent

Question

Kamil Sawicki 2022-5-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1722075-deep-learning-toolbox-commands-sent

回答： Bharathi Yogaraj 2023-6-27

Hello, below I attached some matlab code and Iam wondering what exactly is deployed onto the board and which part of it.

Iam also wondering is Matlab necessary to run further commands after hw.deploy is done or could I use other Interface.

net = resnet18 
hTarget = dlhdl.Target('Xilinx','Interface','Ethernet','IPAddress','192.168.1.101');
hW = dlhdl.Workflow('Network', net, 'Bitstream', 'zcu102_single', 'Target', hTarget)
hW.compile
hW.deploy
I = imread('peppers.png');
I = I(1:224,1:224,1:3);
[prediction, speed] = hW.predict(single(I),'Profile','on');
label = classify(net, I);
figure;
imshow(I);
text(10,20,char(label),'Color','white')

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Bharathi Yogaraj 2023-6-27

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1722075-deep-learning-toolbox-commands-sent#answer_1263418

In above script,

Instructions will be generated for specified network(net) and specified bitstream('zcu102_single') using hW.compile.
Those instructions will be deployed on external DDR and DL IP will be configured in hW.deploy stage.
hW.predict command provides input image for network execution on FPGA and captures output data along with fps information.

For second question, After hW.deploy command, if you would like to feed input from another source/way , should implement system integration interfaces as specified in this documnetation page: https://www.mathworks.com/help/deep-learning-hdl/ug/interface-with-the-deep-learning-processor-ip-core.html.

Here is another example on system integration: https://www.mathworks.com/help/deep-learning-hdl/ug/deploy-and-verify-yolo-v2-vehicle-detector-on-fpga.html