TensorFlow过滤垃圾邮件 tensorflow detection

转载

mob64ca1405a060 2024-04-27 22:39:13

文章标签 TensorFlow过滤垃圾邮件 tensorflow python API 文章分类 机器学习人工智能

最近工作需要折腾 tensorflow，学会图片分类后开始折腾物体检测。感谢前人的工作，现在把物体检跑起来还是比较容易的，但优化就不容易了。

Tensorflow在更新1.2版本之后多了很多新功能，其中放出了很多用tf框架写的深度网络slim模型，大大降低了开发难度，无论是fine-tuning还是修改网络结构都方便了不少。这里讲的的是物体检测（object detection）API，这个库的说明文档很详细。

这个物体检测API提供了5种网络结构的预训练的weights，全部是用COCO数据集进行训练，可以在这里下载：分别是SSD+mobilenet, SSD+inception_v2, R-FCN+resnet101, faster RCNN+resnet101, faster RCNN+inception+resnet101。各个模型的精度和计算所需时间如下，具体测评细节可以看这篇文章：

TensorFlow过滤垃圾邮件 tensorflow detection_tensorflow

依赖包

Protobuf 2.6
Pillow 1.0
lxml
tf Slim 
Jupyter notebook
Matplotlib  # 用这个画图会比较慢，内存占用高，可以用cv2来代替
Tensorflow
Protobuf 2.6
Pillow 1.0
lxml
tf Slim 
Jupyter notebook
Matplotlib  # 用这个画图会比较慢，内存占用高，可以用cv2来代替
Tensorflow

API安装

$ pip install tensorflow-gpu
$ sudo apt-get install protobuf-compiler python-pil python-lxml
$ sudo pip install jupyter
$ sudo pip install matplotlib

因为使用protobuf来配置模型和训练参数，所以API正常使用必须先编译protobuf库：

$ cd tensorflow/models
$ protoc object_detection/protos/*.proto --python_out=.

然后将models和slim（tf高级框架）加入python环境变量：

export PYTHONPATH=$PYTHONPATH:/your/path/to/tensorflow/models:/your/path/to/tensorflow/models/slim

最后测试安装：

python object_detection/builders/model_builder_test.py

fine-tuning

准备数据集以Pascal VOC数据集的格式为例：

object_detection/create_pascal_tf_record.py

提供了一个模板，将voc格式的数据保存到

.record

格式

python object_detection/create_pascal_tf_record.py \
--label_map_path=object_detection/data/pascal_label_map.pbtxt \   # 训练物品的品类和id
--data_dir=VOCdevkit --year=VOC2012 --set=train \
--output_path=pascal_train.record
python object_detection/create_pascal_tf_record.py \
--label_map_path=object_detection/data/pascal_label_map.pbtxt \
--data_dir=VOCdevkit --year=VOC2012 --set=val \
--output_path=pascal_val.record
python object_detection/create_pascal_tf_record.py \
--label_map_path=object_detection/data/pascal_label_map.pbtxt \   # 训练物品的品类和id
--data_dir=VOCdevkit --year=VOC2012 --set=train \
--output_path=pascal_train.record
python object_detection/create_pascal_tf_record.py \
--label_map_path=object_detection/data/pascal_label_map.pbtxt \
--data_dir=VOCdevkit --year=VOC2012 --set=val \
--output_path=pascal_val.record

其中--data_dir为训练集的目录。结构同Pascal VOC，如下：

+ VOCdevkit  # +为文件夹
    + JPEGImages
        - 001.jpg  # - 为文件
    + Annotations
        - 001.xml

训练train和eval输入输出数据储存结构为：

+ input
    - label_map.pbtxt file  # 可以在object_detection/data/*.pbtxt找到样例
    - train TFRecord file
    - eval TFRecord file
+ models
    + modelA
        - pipeline config file # 可以在object_detection/samples/configs/*.config下找到样例，定义训练参数和输入数据
        + train  # 保存训练产生的checkpoint文件
        + eval

准备好上述文件后就可以直接调用train文件进行训练

python object_detection/train.py \
--logtostderr \
--pipeline_config_path=/your/path/to/models/modelA/pipeline config file \ 
--train_dir=/your/path/to/models/modelA/train
python object_detection/train.py \
--logtostderr \
--pipeline_config_path=/your/path/to/models/modelA/pipeline config file \ 
--train_dir=/your/path/to/models/modelA/train

评估
在训练开始以后，就可以运行eval来评估模型的效果。不过实际情况是eval模型也需要加载ckpt文件，因此也需要占用不小的显存，而一般训练的时候都会调整batch尽量利用显卡性能，所以想要实时运行train和eval的话需要调整好两者所需的内存。

python object_detection/eval.py \
--logtostderr \
--pipeline_config_path=/your/path/to/models/modelA/pipeline config file \
--checkpoint_dir=/your/path/to/models/modelA/train \
--eval_dir=/your/path/to/models/modelA/eval
python object_detection/eval.py \
--logtostderr \
--pipeline_config_path=/your/path/to/models/modelA/pipeline config file \
--checkpoint_dir=/your/path/to/models/modelA/train \
--eval_dir=/your/path/to/models/modelA/eval

监控
通过tensorboard命令可以在浏览器很轻松的监控训练进程，在浏览器输入

localhost:6006

（默认）即可

tensorboard --logdir=/your/path/to/models/modelA  # 需要包含eval和train目录(.ckpt， .index， .meta， checkpoint, graph.pbtxt文件)
tensorboard --logdir=/your/path/to/models/modelA  # 需要包含eval和train目录(.ckpt， .index， .meta， checkpoint, graph.pbtxt文件)

freeze model

在训练完成后需要将训练产生的最后一组.meta, .index, .ckpt, checkpoint文件。其中meta保存了graph和metadata，ckpt保存了网络的weights。而在生产环境中进行预测的时候是只需要模型和权重，不需要metadata，所以需要将其提出进行freeze操作，将所需的部分放到一个文件，方便之后的调用，也减少模型加载所需的内存。（在下载的预训练模型解压后可以找到4个文件，其中名为frozen_inference_graph.pb的文件就是freeze后产生的模型文件，比weights文件大，但是比weights和meta文件加起来要小不少。）

本来，tensorflow/python/tools/freeze_graph.py提供了freeze model的api，但是需要提供输出的final node names（一般是softmax之类的最后一层的激活函数命名），而object detection api提供提供了预训练好的网络，final node name并不好找，所以object_detection目录下还提供了export_inference_graph.py。

python export_inference_graph.py \
        --input_type image_tensor \
        --pipeline_config_path /your/path/to/models/modelA/pipeline config file \
        --checkpoint_path  /your/path/to/models/modelA/train/model.ckpt-* \
        --inference_graph_path /your/path/to/models/modelA/train/frozen_inference_graph.pb  # 输出的文件名
python export_inference_graph.py \
        --input_type image_tensor \
        --pipeline_config_path /your/path/to/models/modelA/pipeline config file \
        --checkpoint_path  /your/path/to/models/modelA/train/model.ckpt-* \
        --inference_graph_path /your/path/to/models/modelA/train/frozen_inference_graph.pb  # 输出的文件名

模型调用

目录下提供了一个样例。这里只是稍作调整用cv2来显示图像。

也可以直接使用官方提供的https://github.com/tensorflow/models/blob/master/research/object_detection/object_detection_tutorial.ipynb ，使用jupyter notebook测试。

import numpy as np
import os, sys
import tensorflow as tf
import cv2

MODEL_ROOT = "/home/arkenstone/tensorflow/workspace/models"
sys.path.append(MODEL_ROOT)  # 应用和训练的目录在不同的地方

from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util

MODEL_PATH = "/home/arkenstone/tensorflow/workspace/models/objectdetection/models/faster_rcnn_inception_resnet_v2_atrous_coco_11_06_2017"
PATH_TO_CKPT = MODEL_PATH + '/frozen_inference_graph.pb'  # frozen model path
PATH_TO_LABELS = os.path.join(MODEL_ROOT, 'object_detection/data', 'mscoco_label_map.pbtxt')
NUM_CLASSES = 90

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)  # 格式为{1:{'id': 1, 'name': 'person'}, 2: {'id': 2, 'name': 'bicycle'}, ...}

# 模型加载：test.py
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='')

gpu_memory_fraction = 0.4
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=gpu_memory_fraction)
config = tf.ConfigProto(gpu_options=gpu_options, log_device_placement=False, allow_soft_placement=True)
config.gpu_options.allow_growth = False

def detect(image_path):
    with detection_graph.as_default():  # 需要手动close sess
      with tf.Session(graph=detection_graph, config=config) as sess:
          image = cv2.imread(image_path)
          image_np = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
          image_np_expanded = np.expand_dims(image_np, axis=0)
          image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
          boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
          scores = detection_graph.get_tensor_by_name('detection_scores:0')
          classes = detection_graph.get_tensor_by_name('detection_classes:0')
          num_detections = detection_graph.get_tensor_by_name('num_detections:0')
          (boxes, scores, classes, num_detections) = sess.run(
              [boxes, scores, classes, num_detections],
              feed_dict={image_tensor: image_np_expanded})
          vis_util.visualize_boxes_and_labels_on_image_array(
              image_np,
              np.squeeze(boxes),
              np.squeeze(classes).astype(np.int32),
              np.squeeze(scores),
              category_index,
              use_normalized_coordinates=True,
              line_thickness=4)
          new_img = cv2.cvtColor(image_np, cv2.COLOR_RGB2BGR)
          cv2.imshow("test", new_img)
          cv2.waitKey(0)

if __name__ == '__main__':
    detect(/your/test/image)
import numpy as np
import os, sys
import tensorflow as tf
import cv2

MODEL_ROOT = "/home/arkenstone/tensorflow/workspace/models"
sys.path.append(MODEL_ROOT)  # 应用和训练的目录在不同的地方

from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util

MODEL_PATH = "/home/arkenstone/tensorflow/workspace/models/objectdetection/models/faster_rcnn_inception_resnet_v2_atrous_coco_11_06_2017"
PATH_TO_CKPT = MODEL_PATH + '/frozen_inference_graph.pb'  # frozen model path
PATH_TO_LABELS = os.path.join(MODEL_ROOT, 'object_detection/data', 'mscoco_label_map.pbtxt')
NUM_CLASSES = 90

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)  # 格式为{1:{'id': 1, 'name': 'person'}, 2: {'id': 2, 'name': 'bicycle'}, ...}

# 模型加载：test.py
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='')

gpu_memory_fraction = 0.4
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=gpu_memory_fraction)
config = tf.ConfigProto(gpu_options=gpu_options, log_device_placement=False, allow_soft_placement=True)
config.gpu_options.allow_growth = False

def detect(image_path):
    with detection_graph.as_default():  # 需要手动close sess
      with tf.Session(graph=detection_graph, config=config) as sess:
          image = cv2.imread(image_path)
          image_np = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
          image_np_expanded = np.expand_dims(image_np, axis=0)
          image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
          boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
          scores = detection_graph.get_tensor_by_name('detection_scores:0')
          classes = detection_graph.get_tensor_by_name('detection_classes:0')
          num_detections = detection_graph.get_tensor_by_name('num_detections:0')
          (boxes, scores, classes, num_detections) = sess.run(
              [boxes, scores, classes, num_detections],
              feed_dict={image_tensor: image_np_expanded})
          vis_util.visualize_boxes_and_labels_on_image_array(
              image_np,
              np.squeeze(boxes),
              np.squeeze(classes).astype(np.int32),
              np.squeeze(scores),
              category_index,
              use_normalized_coordinates=True,
              line_thickness=4)
          new_img = cv2.cvtColor(image_np, cv2.COLOR_RGB2BGR)
          cv2.imshow("test", new_img)
          cv2.waitKey(0)

if __name__ == '__main__':
    detect(/your/test/image)