TF2 Detection API Models on mobile TFLite_Detection_PostProcess

1. The entire URL of the documentation with the issue

Running TF2 Detection API Models on mobile

2. Describe the issue

Have previously been able convert and quantize SSD mobilenet following the instructions for TF1 in Running on mobile with TensorFlow Lite. Now I try the same with TF2 but encounter problems running the resulting tflite file. In my case my final hardware requires that the output tensors are TFLite_Detection_PostProcess (they actually can be float32). However, when performing post-training quantization, with the Object detection API, the final tflite output tensors are StatefulPartitonedCall, which are not accepted by the application running on the mcu. So somehow I need to convert to TFLite_Detection_PostProcess.

With TF1 the input_arrays and output_arrays were explicitly defined in the call: bazel run -c opt tensorflow/lite/python:tflite_convert I cant find corresponding information in the present documentation in Running TF2 Detection API Models on mobile.

Can you please update the page with information or some notes for us that are trying to move from TF1 Lite to TF2 Lite?

Thanks
A.