Add TFLite Metadata to TFLite and Edge TPU models (#9903)

export.py

@@ -45,6 +45,7 @@ TensorFlow.js:
 """
 
 import argparse
+import contextlib
 import json
 import os
 import platform
@@ -453,6 +454,39 @@ def export_tfjs(file, prefix=colorstr('TensorFlow.js:')):
     return f, None
 
 
+def add_tflite_metadata(file, metadata, num_outputs):
+    # Add metadata to *.tflite models per https://www.tensorflow.org/lite/models/convert/metadata
+    with contextlib.suppress(ImportError):
+        # check_requirements('tflite_support')
+        from tflite_support import flatbuffers
+        from tflite_support import metadata as _metadata
+        from tflite_support import metadata_schema_py_generated as _metadata_fb
+
+        tmp_file = Path('/tmp/meta.txt')
+        with open(tmp_file, 'w') as meta_f:
+            meta_f.write(str(metadata))
+
+        model_meta = _metadata_fb.ModelMetadataT()
+        label_file = _metadata_fb.AssociatedFileT()
+        label_file.name = tmp_file.name
+        model_meta.associatedFiles = [label_file]
+
+        subgraph = _metadata_fb.SubGraphMetadataT()
+        subgraph.inputTensorMetadata = [_metadata_fb.TensorMetadataT()]
+        subgraph.outputTensorMetadata = [_metadata_fb.TensorMetadataT()] * num_outputs
+        model_meta.subgraphMetadata = [subgraph]
+
+        b = flatbuffers.Builder(0)
+        b.Finish(model_meta.Pack(b), _metadata.MetadataPopulator.METADATA_FILE_IDENTIFIER)
+        metadata_buf = b.Output()
+
+        populator = _metadata.MetadataPopulator.with_model_file(file)
+        populator.load_metadata_buffer(metadata_buf)
+        populator.load_associated_files([str(tmp_file)])
+        populator.populate()
+        tmp_file.unlink()
+
+
 @smart_inference_mode()
 def run(
         data=ROOT / 'data/coco128.yaml',  # 'dataset.yaml path'
@@ -552,6 +586,7 @@ def run(
             f[7], _ = export_tflite(s_model, im, file, int8 or edgetpu, data=data, nms=nms, agnostic_nms=agnostic_nms)
             if edgetpu:
                 f[8], _ = export_edgetpu(file)
+            add_tflite_metadata(f[8] or f[7], metadata, num_outputs=len(s_model.outputs))
         if tfjs:
             f[9], _ = export_tfjs(file)
     if paddle:  # PaddlePaddle

models/common.py

@@ -3,10 +3,13 @@
 Common modules
 """
 
+import ast
+import contextlib
 import json
 import math
 import platform
 import warnings
+import zipfile
 from collections import OrderedDict, namedtuple
 from copy import copy
 from pathlib import Path
@@ -462,6 +465,12 @@ class DetectMultiBackend(nn.Module):
             interpreter.allocate_tensors()  # allocate
             input_details = interpreter.get_input_details()  # inputs
             output_details = interpreter.get_output_details()  # outputs
+            # load metadata
+            with contextlib.suppress(zipfile.BadZipFile):
+                with zipfile.ZipFile(w, "r") as model:
+                    meta_file = model.namelist()[0]
+                    meta = ast.literal_eval(model.read(meta_file).decode("utf-8"))
+                    stride, names = int(meta['stride']), meta['names']
         elif tfjs:  # TF.js
             raise NotImplementedError('ERROR: YOLOv5 TF.js inference is not supported')
         elif paddle:  # PaddlePaddle