tensorflow-lite - TensorFlow Lite Micro 无法从操作码 ADD 获取注册

Question

我正在尝试使用TensorFlow Lite for Microcontrollers在NXP i.MXRT1170上运行此处https://www.tensorflow.org/lite/examples/object_detection/overview提供的对象检测模型。

NXP 提供了图像分类的示例，并且基于该示例我想执行对象检测。因此，第一步只是将执行图像分类的模型更新为执行对象检测的模型并运行示例应用程序并查看应用程序的输出。更换型号后，i.MXRT1170输出错误如：

Didn't find op for builtin opcode 'CONCATENATION' version '1'
Failed to get registration from op code ADD
Failed starting model allocation.
AllocateTensors() failed
Failed initializing model

或者

Didn't find op for builtin opcode 'LOGISTIC' version '1'
Failed to get registration from op code ADD
Failed starting model allocation.
AllocateTensors() failed
Failed initializing model

并且通过向解析器添加所需的操作来解决它们非常简单：

s_microOpResolver.AddConcatenation();
s_microOpResolver.AddLogistic();

但是仍然有如下错误：

Failed to get registration from op code ADD
Failed starting model allocation.
AllocateTensors() failed
Failed initializing model

添加s_microOpResolver.AddAdd();并不能解决问题

有什么想法吗？

标签： tensorflow-lite