Error deploying DNN "Unable to resolve the name coder.inte​rnal.loadD​eepLearnin​gNetwork."

Hello,

I am having an error after compiling an app that content a compiled mex file. I compile a dnn function to GPU mex to increase speed of the deep neural network. Then I compile the application to .exe.

function [categ, score] = classifydnn(resized_cropped_img)
persistent net;

if isempty(net)
    net = coder.loadDeepLearningNetwork('resnet50_0830.mat','net');
end
[categ, score] = classify(net, resized_cropped_img,'ExecutionEnvironment','gpu');
end

But I am having the error

"Unable to resolve the name coder.internal.loadDeepLearningNetwork."

How can I solve this, what I am forgetting to do?

Thanks

Answer 1

[Shashank Gupta]

Hey Wilmer,

The error looks more like an installation package problem, Have you downloaded the required GPU coder packages needed for this function to work. Type "ver" in command line of MATLAB, it will give you the toolboxes you have installed, look for the required toolbox for this function to work. I suspect this has to do with package installation.

Cheers.

[Wilmer Ariza]

Hello,

Thanks for the help

The ver command give me this

-----------------------------------------------------------------------------------------------------
MATLAB Version: 9.9.0.1467703 (R2020b)
MATLAB License Number: 1109990
Operating System: Microsoft Windows 10 Pro Version 10.0 (Build 18363)
Java Version: Java 1.8.0_202-b08 with Oracle Corporation Java HotSpot(TM) 64-Bit Server VM mixed mode
-----------------------------------------------------------------------------------------------------
MATLAB                                                Version 9.9         (R2020b)
Simulink                                              Version 10.2        (R2020b)
Computer Vision Toolbox                               Version 9.3         (R2020b)
Database Toolbox                                      Version 10.0        (R2020b)
Deep Learning Toolbox                                 Version 14.1        (R2020b)
GPU Coder                                             Version 2.0         (R2020b)
Image Acquisition Toolbox                             Version 6.3         (R2020b)
Image Processing Toolbox                              Version 11.2        (R2020b)
MATLAB Coder                                          Version 5.1         (R2020b)
MATLAB Compiler                                       Version 8.1         (R2020b)
Parallel Computing Toolbox                            Version 7.3         (R2020b)
Statistics and Machine Learning Toolbox               Version 12.0        (R2020b)
Symbolic Math Toolbox                                 Version 8.6         (R2020b)

After this I run the setup verification and I get:

gpuEnvObj = coder.gpuEnvConfig;
gpuEnvObj.BasicCodegen = 1;
gpuEnvObj.BasicCodeexec = 1;
gpuEnvObj.DeepLibTarget = 'tensorrt';
gpuEnvObj.DeepCodeexec = 1;
gpuEnvObj.DeepCodegen = 1;
results = coder.checkGpuInstall(gpuEnvObj)
Compatible GPU           : PASSED 
CUDA Environment         : PASSED 
	Runtime   : PASSED 
	cuFFT     : PASSED 
	cuSOLVER  : PASSED 
	cuBLAS    : PASSED 
cuDNN Environment        : PASSED 
TensorRT Environment     : FAILED (Unable to find the 'NVIDIA_TENSORRT' environment variable. Set 'NVIDIA_TENSORRT' to point to the root directory of a TensorRT installation.)
Basic Code Generation    : PASSED 
Basic Code Execution     : PASSED 

results = 

  struct with fields:

                 gpu: 1
                cuda: 1
               cudnn: 1
            tensorrt: 0
        basiccodegen: 1
       basiccodeexec: 1
         deepcodegen: 0
        deepcodeexec: 0
    tensorrtdatatype: 0
           profiling: 0

I dont know if the 0 in deepcodegen is the problem?