Indicatori di correttezza FaceSSD Esempio Colab

Visualizza su TensorFlow.org Esegui in Google Colab Visualizza su GitHub Scarica taccuino

Panoramica

In questa attività, userete Equità indicatori per esplorare le previsioni FaceSSD sui volti etichettate nel set di dati selvaggio . Equità indicatori è una suite di strumenti costruiti in cima tensorflow modello di analisi che consentono di regolare valutazione delle metriche di equità nella pipeline di prodotti.

Informazioni sul set di dati

In questo esercizio, lavorerai con il set di dati di previsione FaceSSD, circa 200k diverse previsioni di immagini e verità di base generate dall'API FaceSSD.

Informazioni sugli strumenti

Tensorflow modello di analisi è una libreria per la valutazione sia per modelli di apprendimento automatico non tensorflow tensorflow e. Consente agli utenti di valutare i propri modelli su grandi quantità di dati in modo distribuito, calcolando in-graph e altre metriche su diverse sezioni di dati e visualizzandoli in notebook.

Tensorflow convalida dei dati è uno strumento che può essere utilizzato per analizzare i dati. Puoi usarlo per trovare potenziali problemi nei tuoi dati, come valori mancanti e squilibri di dati, che possono portare a disparità di equità.

Con Equità indicatori , gli utenti saranno in grado di:

  • Valutare le prestazioni del modello, suddivise in gruppi definiti di utenti
  • Sentiti sicuro dei risultati con intervalli di confidenza e valutazioni a più soglie

importazione

Eseguire il codice seguente per installare la libreria fairness_indicators. Questo pacchetto contiene gli strumenti che utilizzeremo in questo esercizio. Restart Runtime può essere richiesto ma non è necessario.

pip install -q -U pip==20.2
pip install fairness-indicators
import os
import tempfile
import apache_beam as beam
import numpy as np
import pandas as pd
from datetime import datetime

import tensorflow_hub as hub
import tensorflow as tf
import tensorflow_model_analysis as tfma
import tensorflow_data_validation as tfdv
from tensorflow_model_analysis.addons.fairness.post_export_metrics import fairness_indicators
from tensorflow_model_analysis.addons.fairness.view import widget_view
from tensorflow_model_analysis.model_agnostic_eval import model_agnostic_predict as agnostic_predict
from tensorflow_model_analysis.model_agnostic_eval import model_agnostic_evaluate_graph
from tensorflow_model_analysis.model_agnostic_eval import model_agnostic_extractor

from witwidget.notebook.visualization import WitConfigBuilder
from witwidget.notebook.visualization import WitWidget

Scarica e comprendi i dati

Faces etichettati nel selvaggio è un set di dati di riferimento per la verifica pubblica viso, noto anche come coppia corrispondente. LFW contiene più di 13.000 immagini di volti raccolti dal web.

Abbiamo eseguito le previsioni FaceSSD su questo set di dati per prevedere se un volto è presente in una determinata immagine. In questo Colab, divideremo i dati in base al genere per osservare se ci sono differenze significative tra le prestazioni del modello per i diversi gruppi di genere.

Se c'è più di un volto in un'immagine, il sesso è etichettato come "MANCANTE".

Abbiamo ospitato il set di dati su Google Cloud Platform per comodità. Esegui il codice seguente per scaricare i dati da GCP, i dati impiegheranno circa un minuto per essere scaricati e analizzati.

data_location = tf.keras.utils.get_file('lfw_dataset.tf', 'https://storage.googleapis.com/facessd_dataset/lfw_dataset.tfrecord')

stats = tfdv.generate_statistics_from_tfrecord(data_location=data_location)
tfdv.visualize_statistics(stats)
Downloading data from https://storage.googleapis.com/facessd_dataset/lfw_dataset.tfrecord
200835072/200828483 [==============================] - 1s 0us/step
WARNING:apache_beam.runners.interactive.interactive_environment:Dependencies required for Interactive Beam PCollection visualization are not available, please use: `pip install apache-beam[interactive]` to install necessary dependencies to enable all data visualization features.
WARNING:apache_beam.io.tfrecordio:Couldn't find python-snappy so the implementation of _TFRecordUtil._masked_crc32c is not as fast as it could be.
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/tensorflow_data_validation/utils/stats_util.py:247: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.
Instructions for updating:
Use eager execution and: 
`tf.data.TFRecordDataset(path)`
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/tensorflow_data_validation/utils/stats_util.py:247: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.
Instructions for updating:
Use eager execution and: 
`tf.data.TFRecordDataset(path)`

Definizione di costanti

BASE_DIR = tempfile.gettempdir()

tfma_eval_result_path = os.path.join(BASE_DIR, 'tfma_eval_result')

compute_confidence_intervals = True

slice_key = 'object/groundtruth/Gender'
label_key = 'object/groundtruth/face'
prediction_key = 'object/prediction/face'

feature_map = {
    slice_key:
        tf.io.FixedLenFeature([], tf.string, default_value=['none']),
    label_key:
        tf.io.FixedLenFeature([], tf.float32, default_value=[0.0]),
    prediction_key:
        tf.io.FixedLenFeature([], tf.float32, default_value=[0.0]),
}

Configurazione indipendente dal modello per TFMA

model_agnostic_config = agnostic_predict.ModelAgnosticConfig(
    label_keys=[label_key],
    prediction_keys=[prediction_key],
    feature_spec=feature_map)

model_agnostic_extractors = [
    model_agnostic_extractor.ModelAgnosticExtractor(
        model_agnostic_config=model_agnostic_config, desired_batch_size=3),
    tfma.extractors.slice_key_extractor.SliceKeyExtractor(
          [tfma.slicer.SingleSliceSpec(),
           tfma.slicer.SingleSliceSpec(columns=[slice_key])])
]

Richiami di equità e metriche di equità di calcolo

# Helper class for counting examples in beam PCollection
class CountExamples(beam.CombineFn):
    def __init__(self, message):
      self.message = message

    def create_accumulator(self):
      return 0

    def add_input(self, current_sum, element):
      return current_sum + 1

    def merge_accumulators(self, accumulators): 
      return sum(accumulators)

    def extract_output(self, final_sum):
      if final_sum:
        print("%s: %d"%(self.message, final_sum))
metrics_callbacks = [
  tfma.post_export_metrics.fairness_indicators(
      thresholds=[0.1, 0.3, 0.5, 0.7, 0.9],
      labels_key=label_key,
      target_prediction_keys=[prediction_key]),
  tfma.post_export_metrics.auc(
      curve='PR',
      labels_key=label_key,
      target_prediction_keys=[prediction_key]),
]

eval_shared_model = tfma.types.EvalSharedModel(
    add_metrics_callbacks=metrics_callbacks,
    construct_fn=model_agnostic_evaluate_graph.make_construct_fn(
        add_metrics_callbacks=metrics_callbacks,
        config=model_agnostic_config))

with beam.Pipeline() as pipeline:
  # Read data.
  data = (
      pipeline
      | 'ReadData' >> beam.io.ReadFromTFRecord(data_location))

  # Count all examples.
  data_count = (
      data | 'Count number of examples' >> beam.CombineGlobally(
          CountExamples('Before filtering "Gender:MISSING"')))

  # If there are more than one face in image, the gender feature is 'MISSING'
  # and we are filtering that image out.
  def filter_missing_gender(element):
    example = tf.train.Example.FromString(element)
    if example.features.feature[slice_key].bytes_list.value[0] != b'MISSING':
      yield element

  filtered_data = (
      data
      | 'Filter Missing Gender' >> beam.ParDo(filter_missing_gender))

  # Count after filtering "Gender:MISSING".
  filtered_data_count = (
      filtered_data | 'Count number of examples after filtering'
      >> beam.CombineGlobally(
          CountExamples('After filtering "Gender:MISSING"')))

  # Because LFW data set has always faces by default, we are adding
  # labels as 1.0 for all images.
  def add_face_groundtruth(element):
    example = tf.train.Example.FromString(element)
    example.features.feature[label_key].float_list.value[:] = [1.0]
    yield example.SerializeToString()

  final_data = (
      filtered_data
      | 'Add Face Groundtruth' >> beam.ParDo(add_face_groundtruth))

  # Run TFMA.
  _ = (
      final_data
      | 'ExtractEvaluateAndWriteResults' >>
       tfma.ExtractEvaluateAndWriteResults(
                 eval_shared_model=eval_shared_model,
                 compute_confidence_intervals=compute_confidence_intervals,
                 output_path=tfma_eval_result_path,
                 extractors=model_agnostic_extractors))

eval_result = tfma.load_eval_result(output_path=tfma_eval_result_path)
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring send_type hint: <class 'NoneType'>
WARNING:apache_beam.typehints.typehints:Ignoring return_type hint: <class 'NoneType'>
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/tensorflow_model_analysis/post_export_metrics/post_export_metrics.py:178: auc (from tensorflow.python.ops.metrics_impl) is deprecated and will be removed in a future version.
Instructions for updating:
The value of AUC returned by this may race with the update so this is deprecated. Please use tf.keras.metrics.AUC instead.
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.6/site-packages/tensorflow_model_analysis/post_export_metrics/post_export_metrics.py:178: auc (from tensorflow.python.ops.metrics_impl) is deprecated and will be removed in a future version.
Instructions for updating:
The value of AUC returned by this may race with the update so this is deprecated. Please use tf.keras.metrics.AUC instead.
Before filtering "Gender:MISSING": 13836
After filtering "Gender:MISSING": 11544
WARNING:apache_beam.io.filebasedsink:Deleting 1 existing files in target path matching: 
WARNING:apache_beam.io.filebasedsink:Deleting 1 existing files in target path matching: 
WARNING:apache_beam.io.filebasedsink:Deleting 1 existing files in target path matching:

Indicatori di equità di rendering

Visualizza il widget Indicatori di equità con i risultati della valutazione esportati.

Sotto vedrai grafici a barre che mostrano le prestazioni di ogni sezione dei dati sulle metriche selezionate. È possibile regolare la sezione di confronto della linea di base e le soglie visualizzate utilizzando i menu a discesa nella parte superiore della visualizzazione.

Una metrica rilevante per questo caso d'uso è il tasso di vero positivo, noto anche come richiamo. Usa il selettore a sinistra per scegliere il grafico per true_positive_rate. Questi valori di metrica corrispondano ai valori visualizzati sulla carta del modello .

Per alcune foto, il genere è etichettato come giovane invece di maschio o femmina, se la persona nella foto è troppo giovane per essere annotata accuratamente.

widget_view.render_fairness_indicator(eval_result=eval_result,
                                      slicing_column=slice_key)
FairnessIndicatorViewer(slicingMetrics=[{'sliceValue': 'Overall', 'slice': 'Overall', 'metrics': {'post_export…