I am relatively new to object detection using tensorflow and need guidance on the below issue.
I am building a custom model to detect two objects using tensorflow and Faster_Rcnn_inception_v2 model. For this I have used 600 images which contains both the objects. These images are divided into 75% train and 25% test folders. I am able to train with the model on GPU (Linux) machine and achieved loss of only 0.05 After generating frozen_inference_graph.pb file, when I tested, it is not even detecting a single object in over 10 images. It is only working when I lowered the value of min_score_thresh parameter to 0.4 The objects are detected with around 47% confidence.
However, when I trained the same model on different CPU (Windows) machine, it works absolutely fine and results are satisfactory with confidence level above 80 percent.
Can someone please throw some light on this issue? Why the model is not working when trained on GPU but the same model working on CPU?
Note: The issue is occurring only recently, 2 months back, the GPU model was working exceptionally for a different object.
I can share the content of config labelmap or any other file if required.
Command for Training:
python train.py --logtostderr --train_dir="TrainingDp" --pipeline_config_path="TrainingDp/faster_rcnn.config"
Code:
#!/usr/bin/env python
# coding: utf-8
# In[3]:
import os
import pathlib
if "models" in pathlib.Path.cwd().parts:
while "models" in pathlib.Path.cwd().parts:
os.chdir('..')
elif not pathlib.Path('models').exists():
get_ipython().system('git clone --depth 1 https://github.com/tensorflow/models')
# In[4]:
import numpy as np
import os
import six.moves.urllib as urllib
import sys
import tarfile
import tensorflow as tf
import zipfile
import cv2
from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image
from IPython.display import display
# In[5]:
from object_detection.utils import ops as utils_ops
from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util
# In[6]:
# patch tf1 into `utils.ops`
utils_ops.tf = tf.compat.v1
# Patch the location of gfile
tf.gfile = tf.io.gfile
# In[7]:
# 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 = pathlib.Path('C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/test_images')
TEST_IMAGE_PATHS = sorted(list(PATH_TO_TEST_IMAGES_DIR.glob("*.jpg")))
len(TEST_IMAGE_PATHS)
num=len(TEST_IMAGE_PATHS)+1
# In[16]:
model_name = r'C:\Users\xxxxxx\Desktop\models-master\models-master\research\object_detection\TrainingDp2'
PATH_TO_FROZEN_GRAPH= r'C:\Users\xxxxxx\Desktop\models-master\models-master\research\object_detection\inference_graph\frozen_inference_graph.pb'
PATH_TO_LABELS= r'C:\Users\xxxxxx\Desktop\models-master\models-master\research\object_detection\TrainingDp2\labelmap.pbtxt'
# In[17]:
detection_graph = tf.Graph()
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(PATH_TO_FROZEN_GRAPH, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
# In[18]:
category_index = label_map_util.create_category_index_from_labelmap(PATH_TO_LABELS, use_display_name=True)
# In[19]:
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)
# In[20]:
# 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, 10) ]
# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)
# In[21]:
def run_inference_for_single_image(image, graph):
with graph.as_default():
with tf.Session() as sess:
# Get handles to input and output tensors
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:
# The following processing is only for single image
detection_boxes = tf.squeeze(tensor_dict['detection_boxes'], [0])
detection_masks = tf.squeeze(tensor_dict['detection_masks'], [0])
# Reframe is required to translate mask from box coordinates to image coordinates and fit the image size.
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[1], image.shape[2])
detection_masks_reframed = tf.cast(
tf.greater(detection_masks_reframed, 0.5), tf.uint8)
# Follow the convention by adding back the batch dimension
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: image})
# all outputs are float32 numpy arrays, so convert types as appropriate
output_dict['num_detections'] = int(output_dict['num_detections'][0])
output_dict['detection_classes'] = output_dict[
'detection_classes'][0].astype(np.int64)
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
# In[22]:
get_ipython().run_line_magic('matplotlib', 'inline')
# In[23]:
count =0
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(image)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
# Actual detection.
output_dict = run_inference_for_single_image(image_np_expanded, detection_graph)
# Visualization of the results of a detection.
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=4
,min_score_thresh=.4
)
plt.figure(figsize=IMAGE_SIZE)
plt.imshow(image_np)
#cv2.imshow('img',image_np)
RGB=cv2.cvtColor(image_np,cv2.COLOR_BGR2RGB)
filename=r'C:\Users\xxxxxx\Desktop\models-master\models-master\research\object_detection\validation\'iMAGE'+str(count)+'.jpg'
cv2.imwrite(filename,RGB)
count+=1
Content of Config file:
# Faster R-CNN with Inception v2, configured for Oxford-IIIT Pets 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: 2
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: "C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/faster_rcnn_inception_v2_coco_2018_01_28/model.ckpt"
from_detection_checkpoint: true
load_all_detection_checkpoint_vars: true
# Note: The below line limits the training process to 200K steps, which we
# empirically found to be sufficient enough to train the pets 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: "C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/Train.record"
}
label_map_path: "C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/TrainingDp2/labelmap.pbtxt"
}
eval_config: {
metrics_set: "coco_detection_metrics"
num_examples: 1101
}
eval_input_reader: {
tf_record_input_reader {
input_path: "C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/Test.record"
}
label_map_path: "C:/Users/xxxxxx/Desktop/models-master/models-master/research/object_detection/TrainingDp2/labelmap.pbtxt"
shuffle: false
num_readers: 1
}
Thanks.
