Hello, my problem is as follows: I have a motion recognition model under Jetson nano using an ADXL345 accelerometer. I have implemented via the TensorRT project for nvidia the model and it works for a static data coming from the features of a piece of signal of the model given by edge impulse, I even manage to send the prediction by MQTT to a broker under node-red. Now I’m trying to do the same with dynamic sensor data sent by MQTT to the broker and read by the jetson to run the classifier based on the exemple from edge impulse youtube page (https://www.youtube.com/watch?v=OcfgfTIjwz0&t=695s) . Except I get the following error:
And i juste have these type of error when i build my application with APP_CUSTOM=1 TARGET_JETSON_NANO=1 make -j instead of what i used in the past message as APP_CUSTOM=1 TARGET_JETSON_NANO=1 make -j2 :
@Eoin i have encountered many problem to make a github and synchronize the folder with it. Do you have any work email or something like that where i can share you a .zip of my folder
Ps : if you have one and don’t wanna expose it to don’t get a lot of email from other user can you send me a first request of folder at chakib.dj67@gmail.com ?
@Eoin Hello, since your past message i was able to fix my problem with github, there is the link to access to the project named mouvement_projet that i work with :
This folder include all the library that i used, the tflite folder is zipped due to his uncompatible size with github , and there is a cpp-mqtt zip folder that is use for mqtt . in my jetson it was set on this path for the CFLAGS in the makefile : /home/ivan/cpp-mqtt/paho.mqtt.c
