Code compilation for ARM device (Raspberry PI)

Hello,

at the moment I'm trying to compile code for the Raspberry PI. By configuring the coder for 'arm-compute', I get the following error:

dlcfg = coder.DeepLearningConfig('arm-compute');
dlcfg.ArmArchitecture = 'armv7';
dlcfg.ArmComputeVersion = '18.03';

Error using dlcoder_base.internal.checkSupportedTargetLib
Error using coder.DeepLearningConfig. Target library 'arm-compute' is not supported. Supported values for target
library are cudnn , tensorrt , mkldnn.

Error in coder.internal.DeepLearningConfig

Error in coder.DeepLearningConfig (line 24)
    deepLearningCfg = coder.internal.DeepLearningConfig(varargin{:});

How can I make matlab accept the argument 'arm-compute'?

Answer 1

[Bhaskar Vundurthy]

I am assuming that R2019b is being used to in this scenario since ArmComputeVesion is '18.03'. This error might occur when incorrect library functions are used. The correct workflow for utilizing ARM devices (like Raspberry Pi) can be found here.

Pinpointing the error any further might require knowledge of the actual code being used.