This is the Arduino® compatible port of the AIfES machine learning framework, developed and maintained by Fraunhofer Institute for Microelectronic Circuits and Systems.
After merging #18 the wrong macro is used when initializing the sigmoid layer in the example "0_Universal/6_CNN/1_CNN_Training_F32/". However, this does not affect the result of the training!
After merging #18 the wrong macro is used when initializing the sigmoid layer in the example "0_Universal/6_CNN/1_CNN_Training_F32/". However, this does not affect the result of the training!
In line 141, instead of
ailayer_sigmoid_f32_t sigmoid_layer_3 = AILAYER_RELU_F32_A();
should be
ailayer_sigmoid_f32_t sigmoid_layer_3 = AILAYER_SIGMOID_F32_A();
should be used.
https://github.com/Fraunhofer-IMS/AIfES_for_Arduino/blob/07a438213f67e8f9d0271417a9f16be97f1597d2/examples/0_Universal/6_CNN/1_CNN_Training_F32/1_CNN_Training_F32.ino#L141C2-L141C65