基于TensorFlow Object Detection API进行迁移学习训练自己的人脸检测模型（二） 基于TensorFlow Object Detection API进行迁移学习训练自己的人脸检测模型（一）

前言

已完成数据预处理工作，具体参照：

基于TensorFlow Object Detection API进行迁移学习训练自己的人脸检测模型（一）

设置配置文件

新建目录face_faster_rcnn

将上文已完成预数据处理的目录data移动至face_faster_rcnn目录下，

并在face_faster_rcnn目录下创建face_label.pbtxt文件，内容如下：

item {
  id: 1
  name: 'face'
}

在已下载的TensorFlow Object Detection API目录下搜索faster_rcnn_inception_v2_coco.config，具体目录models-master\research\object_detection\samples\configs，将其拷贝至face_faster_rcnn目录下

在存储库中,faster_rcnn_inception_v2_coco.config文件用来训练人工神经网络的配置文件。该文件基于pet检测器。
在本例中，num_classes的数量仍然是一个，因为只有人脸才会被识别。

变量fine_tune_checkpoint用于指示以前模型的路径以获得学习。微调检查点文件（fine tune checkpoint file）在应用转移学习上被使用。转移学习是一种机器学习方法，它专注于将从一个问题中获得的知识应用到另一个问题上。

在类train_input_reader中，用带有TFRecord文件的链接以训练模型。在配置文件中，需要将其自定义到正确的位置。
变量label_map_path包含索引ID和名称。使用这个文件，0被用作占位符，所以我们从数字1开始。

验证有两个很重要的变量。在eval_config类中的变量 num_examples用于设置示例的数量。
eval_input_reader类描述了验证数据的位置。在这个位置也有一条路径。

此外，还可以改变学习速度、批量大小和其他设置。现在，我保留了默认设置。

完整的faster_rcnn_inception_v2_coco.config

# Faster R-CNN with Inception v2, configuration for MSCOCO Dataset.
# Users should configure the fine_tune_checkpoint field in the train config as
# well as the label_map_path and input_path fields in the train_input_reader and
# eval_input_reader. Search for "PATH_TO_BE_CONFIGURED" to find the fields that
# should be configured.


model {
  faster_rcnn {
    num_classes: 1
    image_resizer {
      keep_aspect_ratio_resizer {
        min_dimension: 600
        max_dimension: 1024
      }
    }
    feature_extractor {
      type: 'faster_rcnn_inception_v2'
      first_stage_features_stride: 16
    }
    first_stage_anchor_generator {
      grid_anchor_generator {
        scales: [0.25, 0.5, 1.0, 2.0]
        aspect_ratios: [0.5, 1.0, 2.0]
        height_stride: 16
        width_stride: 16
      }
    }
    first_stage_box_predictor_conv_hyperparams {
      op: CONV
      regularizer {
        l2_regularizer {
          weight: 0.0
        }
      }
      initializer {
        truncated_normal_initializer {
          stddev: 0.01
        }
      }
    }
    first_stage_nms_score_threshold: 0.0
    first_stage_nms_iou_threshold: 0.7
    first_stage_max_proposals: 300
    first_stage_localization_loss_weight: 2.0
    first_stage_objectness_loss_weight: 1.0
    initial_crop_size: 14
    maxpool_kernel_size: 2
    maxpool_stride: 2
    second_stage_box_predictor {
      mask_rcnn_box_predictor {
        use_dropout: false
        dropout_keep_probability: 1.0
        fc_hyperparams {
          op: FC
          regularizer {
            l2_regularizer {
              weight: 0.0
            }
          }
          initializer {
            variance_scaling_initializer {
              factor: 1.0
              uniform: true
              mode: FAN_AVG
            }
          }
        }
      }
    }
    second_stage_post_processing {
      batch_non_max_suppression {
        score_threshold: 0.0
        iou_threshold: 0.6
        max_detections_per_class: 100
        max_total_detections: 300
      }
      score_converter: SOFTMAX
    }
    second_stage_localization_loss_weight: 2.0
    second_stage_classification_loss_weight: 1.0
  }
}

train_config: {
  batch_size: 1
  optimizer {
    momentum_optimizer: {
      learning_rate: {
        manual_step_learning_rate {
          initial_learning_rate: 0.0002
          schedule {
            step: 900000
            learning_rate: .00002
          }
          schedule {
            step: 1200000
            learning_rate: .000002
          }
        }
      }
      momentum_optimizer_value: 0.9
    }
    use_moving_average: false
  }
  gradient_clipping_by_norm: 10.0
  fine_tune_checkpoint: "data/faster_rcnn_inception_v2_coco_2018_01_28/model.ckpt"
  from_detection_checkpoint: true
  # Note: The below line limits the training process to 200K steps, which we
  # empirically found to be sufficient enough to train the COCO dataset. This
  # effectively bypasses the learning rate schedule (the learning rate will
  # never decay). Remove the below line to train indefinitely.
  num_steps: 200000
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
}

train_input_reader: {
  tf_record_input_reader {
    input_path: "data/train.record"
  }
  label_map_path: "face_label.pbtxt"
}

eval_config: {
  num_examples: 8000
  # Note: The below line limits the evaluation process to 10 evaluations.
  # Remove the below line to evaluate indefinitely.
  max_evals: 10
}

eval_input_reader: {
  tf_record_input_reader {
    input_path: "data/val.record"
  }
  label_map_path: "face_label.pbtxt"
  shuffle: false
  num_readers: 1
}

训练

将models-master\research目录下的object_detection文件夹整个拷贝至face_faster_rcnn目录下，并将object_detection目录下的train.py、export_inference_graph.py、eval.py三个文件拷贝至face_faster_rcnn目录，至此目录结构全部完成

红框可忽略，为已训练的模型

现在，它将开始真正的工作。计算机将从人脸检测数据集中学习并建立一个神经网络。当我在CPU上模拟训练时，需要几天的时间才能得到一个好的结果。但强大的Nvidia显卡可以将时间缩短为几个小时。

python train.py --logtostderr --train_dir=data/ --pipeline_config_path=faster_rcnn_inception_v2_coco.config  --train_dir=model_output

运行会出现以下异常

ValueError: Tried to convert 't' to a tensor and failed. Error: Argument must be a dense tensor: range(0, 3) - got shape [3], but wanted [].

可能是python3兼容性问题

把object_detection/utils/learning_schedules.py文件的 第167-169行由

# ＃　修改１６７－１６９
rate_index = tf.reduce_max(tf.where(tf.greater_equal(global_step, boundaries),
                                    range(num_boundaries),
                                    [0] * num_boundaries))
# ＃　成
rate_index = tf.reduce_max(tf.where(tf.greater_equal(global_step, boundaries),
                                    list(range(num_boundaries)),
                                    [0] * num_boundaries))

Tensorboard深入的了解了学习过程。该工具是Tensorflow的一部分，且可以自动安装。

tensorboard --logdir=model_output

检查点转换为protobuf变为可运行的模型

在带有计算机视觉库Tensorflow目标识别检测中使用该模型。 以下命令提供了模型存储库的位置和最后一个检查点。文件夹文件夹将包含frozen_inference_graph.pb

python export_inference_graph.py --input_type image_tensor --pipeline_config_path faster_rcnn_inception_v2_coco.config --trained_checkpoint_prefix model_output/model.ckpt-11543 --output_directory face_faster_rcnn_model/

打包模型

tar zcvf face_faster_rcnn_model.tar.gz face_faster_rcnn_model

预测新图片

编写ImageTest.py

# coding: utf-8
import numpy as np
import os
import six.moves.urllib as urllib
import sys
import tarfile
import tensorflow as tf
import zipfile

from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image

# This is needed since the notebook is stored in the object_detection folder.
#sys.path.append("..")
from object_detection.utils import ops as utils_ops

if tf.__version__ < '1.4.0':
    raise ImportError('Please upgrade your tensorflow installation to v1.4.* or later!')
  
# This is needed to display the images.
from object_detection.utils import label_map_util

from object_detection.utils import visualization_utils as vis_util

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
##Model preparation##

# What model to download.
MODEL_NAME = 'faster_rcnn_inception_v2_coco_2018_01_28'
MODEL_FILE = MODEL_NAME + '.tar.gz'
DOWNLOAD_BASE = 'http://download.tensorflow.org/models/object_detection/'

# Path to frozen detection graph. This is the actual model that is used for the object detection.
PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'

# List of the strings that is used to add correct label for each box.
PATH_TO_LABELS = os.path.join('data', 'face_label.pbtxt')

NUM_CLASSES = 1

## Download Model##
#opener = urllib.request.URLopener()
#opener.retrieve(DOWNLOAD_BASE + MODEL_FILE, MODEL_FILE)
tar_file = tarfile.open(MODEL_FILE)
for file in tar_file.getmembers():
  file_name = os.path.basename(file.name)
  if 'frozen_inference_graph.pb' in file_name:
    tar_file.extract(file, os.getcwd())
    
## Load a (frozen) Tensorflow model into memory.
detection_graph = tf.Graph()
with detection_graph.as_default():
  od_graph_def = tf.GraphDef()
  with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
    serialized_graph = fid.read()
    od_graph_def.ParseFromString(serialized_graph)
    tf.import_graph_def(od_graph_def, name='')
    
## Loading label map
label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

def load_image_into_numpy_array(image):
  (im_width, im_height) = image.size
  return np.array(image.getdata()).reshape(
      (im_height, im_width, 3)).astype(np.uint8)
      
# For the sake of simplicity we will use only 2 images:
# image1.jpg
# image2.jpg
# If you want to test the code with your images, just add path to the images to the TEST_IMAGE_PATHS.
PATH_TO_TEST_IMAGES_DIR = 'test_images'
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 3) ]

# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)

def run_inference_for_single_image(image, graph):
  with graph.as_default():
    with tf.Session() as sess:
      # 获取输入和输出张量
      ops = tf.get_default_graph().get_operations()
      all_tensor_names = {output.name for op in ops for output in op.outputs}
      tensor_dict = {}
      for key in [
          'num_detections', 'detection_boxes', 'detection_scores',
          'detection_classes', 'detection_masks'
      ]:
        tensor_name = key + ':0'
        if tensor_name in all_tensor_names:
          tensor_dict[key] = tf.get_default_graph().get_tensor_by_name(
              tensor_name)
      if 'detection_masks' in tensor_dict:
        # 以下处理仅针对单个图像
        detection_boxes = tf.squeeze(tensor_dict['detection_boxes'], [0])
        detection_masks = tf.squeeze(tensor_dict['detection_masks'], [0])
        # 需要重从框坐标转换成图像坐标，并适合图像大小。
        real_num_detection = tf.cast(tensor_dict['num_detections'][0], tf.int32)
        detection_boxes = tf.slice(detection_boxes, [0, 0], [real_num_detection, -1])
        detection_masks = tf.slice(detection_masks, [0, 0, 0], [real_num_detection, -1, -1])
        detection_masks_reframed = utils_ops.reframe_box_masks_to_image_masks(
            detection_masks, detection_boxes, image.shape[0], image.shape[1])
        detection_masks_reframed = tf.cast(
            tf.greater(detection_masks_reframed, 0.5), tf.uint8)
        # 通过添加批次尺寸来遵循惯例
        tensor_dict['detection_masks'] = tf.expand_dims(
            detection_masks_reframed, 0)
      image_tensor = tf.get_default_graph().get_tensor_by_name('image_tensor:0')

      # Run inference
      output_dict = sess.run(tensor_dict,
                             feed_dict={image_tensor: np.expand_dims(image, 0)})

      # 所有输出都是FLUAT32 NUMPY数组，以适应转换类型
      output_dict['num_detections'] = int(output_dict['num_detections'][0])
      output_dict['detection_classes'] = output_dict[
          'detection_classes'][0].astype(np.uint8)
      output_dict['detection_boxes'] = output_dict['detection_boxes'][0]
      output_dict['detection_scores'] = output_dict['detection_scores'][0]
      if 'detection_masks' in output_dict:
        output_dict['detection_masks'] = output_dict['detection_masks'][0]
  return output_dict              

for image_path in TEST_IMAGE_PATHS:
  image = Image.open(image_path)
  # 这个array在之后会被用来准备为图片加上框和标签
  image_np = load_image_into_numpy_array(image)
  # 扩展维度，因为模型期望图像具有形状：: [1, None, None, 3]
  image_np_expanded = np.expand_dims(image_np, axis=0)
  # 执行侦测任务
  output_dict = run_inference_for_single_image(image_np, detection_graph)
  # 检测结果的可视化
  vis_util.visualize_boxes_and_labels_on_image_array(
      image_np,
      output_dict['detection_boxes'],
      output_dict['detection_classes'],
      output_dict['detection_scores'],
      category_index,
      instance_masks=output_dict.get('detection_masks'),
      use_normalized_coordinates=True,
      line_thickness=8)
  plt.figure(figsize=IMAGE_SIZE)
  plt.imshow(image_np)
  plt.show()
  

新建test_images文件夹，并放入图片

修改MODEL_NAME，PATH_TO_LABELS

python ImageTest.py

效果

评估

除了用于Tensorflow目标识别检测训练的数据外，还有一个评估数据集。基于此评估数据集，可以计算精度。对于我的模型，我计算了精度(平均精度)。我以14392步的速度获得了83.80％的分数(epochs)。对于这个过程，Tensorflow有一个脚本，使它可以在Tensorboard中看到分数是多少。除了训练之外，建议你运行评估过程。

python eval.py --logtostderr --pipeline_config_path=faster_rcnn_inception_v2_coco.config  --checkpoint_dir=model_output --eval_dir=eval