page_type: reference description: Writes metadata into an object detector.

tflite_support.metadata_writers.object_detector.MetadataWriter

<table class="tfo-notebook-buttons tfo-api nocontent" align="left"> <td> <a target="_blank" href="https://github.com/tensorflow/tflite-support/blob/v0.4.4/tensorflow_lite_support/metadata/python/metadata_writers/object_detector.py#L103-L295">

View source on GitHub

</a> </td> </table>

Writes metadata into an object detector.

Inherits From: MetadataWriter

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>tflite_support.metadata_writers.object_detector.MetadataWriter( model_buffer: bytearray, metadata_buffer: Optional[bytearray] = None, associated_files: Optional[List[str]] = None ) </code></pre>   <table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2"><h2 class="add-link">Args</h2></th></tr> <tr> <td> `model_buffer`<a id="model_buffer"></a> </td> <td> valid buffer of the model file. </td> </tr><tr> <td> `metadata_buffer`<a id="metadata_buffer"></a> </td> <td> valid buffer of the metadata. </td> </tr><tr> <td> `associated_files`<a id="associated_files"></a> </td> <td> path to the associated files to be populated. </td> </tr> </table>

Methods

<h3 id="create_for_inference"><code>create_for_inference</code></h3>

<a target="_blank" class="external" href="https://github.com/tensorflow/tflite-support/blob/v0.4.4/tensorflow_lite_support/metadata/python/metadata_writers/object_detector.py#L234-L295">View source</a>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>@classmethod</code> <code>create_for_inference( model_buffer: bytearray, input_norm_mean: List[float], input_norm_std: List[float], label_file_paths: List[str], score_calibration_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/ScoreCalibrationMd"><code>tflite_support.metadata_writers.metadata_info.ScoreCalibrationMd</code></a>] = None ) </code></pre>

Creates mandatory metadata for TFLite Support inference.

The parameters required in this method are mandatory when using TFLite Support features, such as Task library and Codegen tool (Android Studio ML Binding). Other metadata fields will be set to default. If other fields need to be filled, use the method create_from_metadata_info to edit them.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Args</th></tr> <tr> <td> `model_buffer` </td> <td> valid buffer of the model file. </td> </tr><tr> <td> `input_norm_mean` </td> <td> the mean value used in the input tensor normalization [1]. </td> </tr><tr> <td> `input_norm_std` </td> <td> the std value used in the input tensor normalizarion [1]. </td> </tr><tr> <td> `label_file_paths` </td> <td> paths to the label files [2] in the category tensor. Pass in an empty list, If the model does not have any label file. </td> </tr><tr> <td> `score_calibration_md` </td> <td> information of the score calibration operation [3] in the classification tensor. Optional if the model does not use score calibration. [1]: https://www.tensorflow.org/lite/convert/metadata#normalization_and_quantization_parameters [2]: https://github.com/tensorflow/tflite-support/blob/b80289c4cd1224d0e1836c7654e82f070f9eefaa/tensorflow_lite_support/metadata/metadata_schema.fbs#L108 [3]: https://github.com/tensorflow/tflite-support/blob/5e0cdf5460788c481f5cd18aab8728ec36cf9733/tensorflow_lite_support/metadata/metadata_schema.fbs#L434 </td> </tr> </table>  <table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> A MetadataWriter object. </td> </tr> </table> <h3 id="create_from_metadata"><code>create_from_metadata</code></h3>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>@classmethod</code> <code>create_from_metadata( model_buffer: bytearray, model_metadata: Optional[<a href="../../../tflite_support/metadata_schema_py_generated/ModelMetadataT"><code>tflite_support.metadata_schema_py_generated.ModelMetadataT</code></a>] = None, input_metadata: Optional[List[_metadata_fb.TensorMetadataT]] = None, output_metadata: Optional[List[_metadata_fb.TensorMetadataT]] = None, associated_files: Optional[List[str]] = None, input_process_units: Optional[List[_metadata_fb.ProcessUnitT]] = None, output_process_units: Optional[List[_metadata_fb.ProcessUnitT]] = None ) </code></pre>

Creates MetadataWriter based on the metadata Flatbuffers Python Objects.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Args</th></tr> <tr> <td> `model_buffer` </td> <td> valid buffer of the model file. </td> </tr><tr> <td> `model_metadata` </td> <td> general model metadata [1]. The subgraph_metadata will be refreshed with input_metadata and output_metadata. </td> </tr><tr> <td> `input_metadata` </td> <td> a list of metadata of the input tensors [2]. </td> </tr><tr> <td> `output_metadata` </td> <td> a list of metadata of the output tensors [3]. </td> </tr><tr> <td> `associated_files` </td> <td> path to the associated files to be populated. </td> </tr><tr> <td> `input_process_units` </td> <td> a lits of metadata of the input process units [4]. </td> </tr><tr> <td> `output_process_units` </td> <td> a lits of metadata of the output process units [5]. [1]: https://github.com/tensorflow/tflite-support/blob/b80289c4cd1224d0e1836c7654e82f070f9eefaa/tensorflow_lite_support/metadata/metadata_schema.fbs#L640-L681 [2]: https://github.com/tensorflow/tflite-support/blob/b80289c4cd1224d0e1836c7654e82f070f9eefaa/tensorflow_lite_support/metadata/metadata_schema.fbs#L590 [3]: https://github.com/tensorflow/tflite-support/blob/b80289c4cd1224d0e1836c7654e82f070f9eefaa/tensorflow_lite_support/metadata/metadata_schema.fbs#L599 [4]: https://github.com/tensorflow/tflite-support/blob/b5cc57c74f7990d8bc055795dfe8d50267064a57/tensorflow_lite_support/metadata/metadata_schema.fbs#L646 [5]: https://github.com/tensorflow/tflite-support/blob/b5cc57c74f7990d8bc055795dfe8d50267064a57/tensorflow_lite_support/metadata/metadata_schema.fbs#L650 </td> </tr> </table>  <table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> A MetadataWriter Object. </td> </tr> </table> <h3 id="create_from_metadata_info"><code>create_from_metadata_info</code></h3>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>@classmethod</code> <code>create_from_metadata_info( model_buffer: bytearray, general_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/GeneralMd"><code>tflite_support.metadata_writers.metadata_info.GeneralMd</code></a>] = None, input_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/InputImageTensorMd"><code>tflite_support.metadata_writers.metadata_info.InputImageTensorMd</code></a>] = None, output_location_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/TensorMd"><code>tflite_support.metadata_writers.metadata_info.TensorMd</code></a>] = None, output_category_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/CategoryTensorMd"><code>tflite_support.metadata_writers.metadata_info.CategoryTensorMd</code></a>] = None, output_score_md: Union[None, <a href="../../../tflite_support/metadata_writers/metadata_info/TensorMd"><code>tflite_support.metadata_writers.metadata_info.TensorMd</code></a>, <a href="../../../tflite_support/metadata_writers/metadata_info/ClassificationTensorMd"><code>tflite_support.metadata_writers.metadata_info.ClassificationTensorMd</code></a>] = None, output_number_md: Optional[<a href="../../../tflite_support/metadata_writers/metadata_info/TensorMd"><code>tflite_support.metadata_writers.metadata_info.TensorMd</code></a>] = None ) </code></pre>

Creates MetadataWriter based on general/input/outputs information.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Args</th></tr> <tr> <td> `model_buffer` </td> <td> valid buffer of the model file. </td> </tr><tr> <td> `general_md` </td> <td> general information about the model. </td> </tr><tr> <td> `input_md` </td> <td> input image tensor informaton. </td> </tr><tr> <td> `output_location_md` </td> <td> output location tensor informaton. The location tensor is a multidimensional array of [N][4] floating point values between 0 and 1, the inner arrays representing bounding boxes in the form [top, left, bottom, right]. </td> </tr><tr> <td> `output_category_md` </td> <td> output category tensor information. The category tensor is an array of N integers (output as floating point values) each indicating the index of a class label from the labels file. </td> </tr><tr> <td> `output_score_md` </td> <td> output score tensor information. The score tensor is an array of N floating point values between 0 and 1 representing probability that a class was detected. Use ClassificationTensorMd to calibrate score. </td> </tr><tr> <td> `output_number_md` </td> <td> output number of detections tensor information. This tensor is an integer value of N. </td> </tr> </table>  <table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> A MetadataWriter object. </td> </tr> </table> <h3 id="get_metadata_json"><code>get_metadata_json</code></h3>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>get_metadata_json() -> str </code></pre>

Gets the generated JSON metadata string before populated into model.

This method returns the metadata buffer before populated into the model. More fields could be filled by MetadataPopulator, such as min_parser_version. Use get_populated_metadata_json() if you want to get the final metadata string.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> The generated JSON metadata string before populated into model. </td> </tr> </table> <h3 id="get_populated_metadata_json"><code>get_populated_metadata_json</code></h3>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>get_populated_metadata_json() -> str </code></pre>

Gets the generated JSON metadata string after populated into model.

More fields could be filled by MetadataPopulator, such as min_parser_version. Use get_metadata_json() if you want to get the original metadata string.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> The generated JSON metadata string after populated into model. </td> </tr> </table> <h3 id="populate"><code>populate</code></h3>

<pre class="devsite-click-to-copy prettyprint lang-py tfo-signature-link"> <code>populate() -> bytearray </code></pre>

Populates the metadata and label file to the model file.

<table class="responsive fixed orange"> <colgroup><col width="214px"><col></colgroup> <tr><th colspan="2">Returns</th></tr> <tr class="alt"> <td colspan="2"> A new model buffer with the metadata and associated files. </td> </tr> </table>