def main(argv):
flags = parser(
description="freeze yolov3 graph from checkpoint file").parse_args()
classes = utils.read_coco_names("./data/coco.names")
num_classes = len(classes)
SIZE = flags.image_size
print("=> the input image size is [%d, %d]" % (SIZE, SIZE))
model = yolov3.yolov3(num_classes)
with tf.Graph().as_default() as graph:
sess = tf.Session(graph=graph)
inputs = tf.placeholder(
tf.float32, [1, SIZE, SIZE, 3]) # placeholder for detector inputs
with tf.variable_scope('yolov3'):
feature_map = model.forward(inputs, is_training=False)
boxes, confs, probs = model.predict(feature_map)
scores = confs * probs
print("=>", boxes, scores)
boxes, scores, labels = utils.gpu_nms(
boxes,
scores,
num_classes,
score_thresh=flags.score_threshold,
iou_thresh=flags.iou_threshold)
print("=>", boxes, scores, labels)
feature_map_1, feature_map_2, feature_map_3 = feature_map
print("=>", feature_map_1, feature_map_2, feature_map_3)
saver = tf.train.Saver(var_list=tf.global_variables(scope='yolov3'))
if flags.convert:
if not os.path.exists(flags.weights_path):
url = 'https://github.com/YunYang1994/tensorflow-yolov3/releases/download/v1.0/yolov3.weights'
for i in range(3):
time.sleep(1)
print("=> %s does not exists ! " % flags.weights_path)
print("=> It will take a while to download it from %s" % url)
print('=> Downloading yolov3 weights ... ')
wget.download(url, flags.weights_path)
load_ops = utils.load_weights(tf.global_variables(scope='yolov3'),
flags.weights_path)
sess.run(load_ops)
save_path = saver.save(sess, save_path=flags.ckpt_file)
print('=> model saved in path: {}'.format(save_path))
if flags.freeze:
saver.restore(sess, flags.ckpt_file)
print('=> checkpoint file restored from ', flags.ckpt_file)
utils.freeze_graph(sess, './checkpoint/yolov3_cpu_nms.pb',
["concat_9", "mul_6"])
utils.freeze_graph(sess, './checkpoint/yolov3_gpu_nms.pb',
["concat_10", "concat_11", "concat_12"])
utils.freeze_graph(sess, './checkpoint/yolov3_feature.pb', [
"yolov3/yolo-v3/feature_map_1",
"yolov3/yolo-v3/feature_map_2",
"yolov3/yolo-v3/feature_map_3",
])
def __init__(self):
''' Called when class object is created. '''
self.IMAGE_H, self.IMAGE_W = imgsize, imgsize
self.classes = utils.read_coco_names('./data/coco.names')
self.num_classes = len(self.classes)
self.colors = self.color_generator()
self.windowName = 'Inference'
cv2.namedWindow(self.windowName, cv2.WINDOW_NORMAL)
def __init__(self):
''' Called when class object is created. '''
self.img_size = imgsize
self.max_batch_size = batchsize
self.num_classes = len(utils.read_coco_names('./data/coco.names'))
self.input_tensor, self.output_tensors = utils.read_pb_return_tensors(
tf.compat.v1.get_default_graph(), "./checkpoint/yolov3_cpu_nms.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
self.config = tf.compat.v1.ConfigProto()
# self.config.gpu_options.per_process_gpu_memory_fraction = 0.4
self.sess = tf.compat.v1.Session(config=self.config)
_ = self.sess.run(
self.output_tensors,
feed_dict={self.input_tensor: self.createRandomSample()})
def __init__(self):
self.SIZE = [416, 416]
self.classes = utils.read_coco_names('./data/coco.names')
self.num_classes = len(self.classes)
self.input_tensor, self.output_tensors = utils.read_pb_return_tensors(
tf.get_default_graph(), "./checkpoint/yolov3_gpu_nms.pb",
["Placeholder:0", "concat_10:0", "concat_11:0", "concat_12:0"])
self.sess = tf.Session()
self.LOWER_RED_RANGE = np.array([17, 15, 100])
self.UPPER_RED_RANGE = np.array([50, 56, 200])
self.pitch_rate = 0
self.yaw_rate = 0
self.vertical_rate = 0
self.TARGET = 0
self.drone_centroid = (int(856 / 2), int(480 * (0.4))
) # drone_centroid
self.result = None
def __init__(self):
self.IMAGE_H = 416
self.IMAGE_W = 416
self.classes = utils.read_coco_names('./data/coco.names')
self.num_classes = len(self.classes)
self.gpu_nms_graph = tf.Graph()
self.input_tensor, self.output_tensors = utils.read_pb_return_tensors(
self.gpu_nms_graph,
"./checkpoint/yolov3_gpu_nms.pb",
["Placeholder:0", "concat_10:0", "concat_11:0", "concat_12:0"])
self.sess = tf.Session(graph=self.gpu_nms_graph)
self.last_PIL_image = None # 原图
self.last_boxes = None
self.last_scores = None
self.last_labels = None
self.colors = [[254.0, 254.0, 254], [248.92, 228.6, 127], [243.84, 203.20000000000002, 0], [238.76, 177.79999999999998, -127], [233.68, 152.4, -254], [228.6, 127.0, 254], [223.52, 101.60000000000001, 127], [218.44, 76.20000000000002, 0], [213.35999999999999, 50.79999999999999, -127], [208.28000000000003, 25.399999999999995, -254], [203.20000000000002, 0.0, 254], [198.12, -25.400000000000023, 127], [193.04, -50.79999999999999, 0], [187.96, -76.20000000000002, -127], [182.88, -101.59999999999998, -254], [177.79999999999998, -127.0, 254], [172.71999999999997, -152.40000000000003, 127], [167.64, -177.79999999999998, 0], [162.56, -203.20000000000002, -127], [157.48, -228.59999999999997, -254], [152.4, -254.0, 254], [147.32000000000002, -279.40000000000003, 127], [142.24, -304.80000000000007, 0], [137.16, -330.19999999999993, -127], [132.08, -355.59999999999997, -254], [127.0, 254.0, 254], [121.92, 228.6, 127], [116.83999999999999, 203.20000000000002, 0], [111.75999999999999, 177.79999999999998, -127], [106.68, 152.4, -254], [101.60000000000001, 127.0, 254], [96.52, 101.60000000000001, 127], [91.44, 76.20000000000002, 0], [86.35999999999999, 50.79999999999999, -127], [81.27999999999999, 25.399999999999995, -254], [76.20000000000002, 0.0, 254], [71.12, -25.400000000000023, 127], [66.04, -50.79999999999999, 0], [60.96, -76.20000000000002, -127], [55.879999999999995, -101.59999999999998, -254], [50.79999999999999, -127.0, 254], [45.72000000000001, -152.40000000000003, 127], [40.64000000000001, -177.79999999999998, 0], [35.56, -203.20000000000002, -127], [30.48, -228.59999999999997, -254], [25.399999999999995, -254.0, 254], [20.31999999999999, -279.40000000000003, 127], [15.240000000000013, -304.80000000000007, 0], [10.160000000000009, -330.19999999999993, -127], [5.0800000000000045, -355.59999999999997, -254], [0.0, 254.0, 254], [-5.0800000000000045, 228.6, 127], [-10.160000000000009, 203.20000000000002, 0], [-15.240000000000013, 177.79999999999998, -127], [-20.320000000000018, 152.4, -254], [-25.400000000000023, 127.0, 254], [-30.480000000000025, 101.60000000000001, 127], [-35.559999999999974, 76.20000000000002, 0], [-40.63999999999998, 50.79999999999999, -127], [-45.719999999999985, 25.399999999999995, -254], [-50.79999999999999, 0.0, 254], [-55.879999999999995, -25.400000000000023, 127], [-60.96, -50.79999999999999, 0], [-66.04, -76.20000000000002, -127], [-71.12, -101.59999999999998, -254], [-76.20000000000002, -127.0, 254], [-81.28000000000002, -152.40000000000003, 127], [-86.36000000000001, -177.79999999999998, 0], [-91.44000000000003, -203.20000000000002, -127], [-96.51999999999997, -228.59999999999997, -254], [-101.59999999999998, -254.0, 254], [-106.67999999999998, -279.40000000000003, 127], [-111.75999999999999, -304.80000000000007, 0], [-116.83999999999999, -330.19999999999993, -127], [-121.92, -355.59999999999997, -254], [-127.0, 254.0, 254], [-132.08, 228.6, 127], [-137.16, 203.20000000000002, 0], [-142.24, 177.79999999999998, -127], [-147.32000000000002, 152.4, -254]]
self.label_name = ['person', 'bicycle', 'car', 'motorbike', 'aeroplane', 'bus', 'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', 'parking meter',
'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', 'sports ball',
'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange',
'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'sofa', 'pottedplant', 'bed', 'diningtable', 'toilet', 'tvmonitor', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven',
'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', 'hair drier', 'toothbrush']
# Created date: 2018-11-30 15:47:45
# Description :
#
#================================================================
import tensorflow as tf
from core import utils, yolov3
from core.dataset import dataset, Parser
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 8
EPOCHS = 2000 * 1000
LR = 0.0001
SHUFFLE_SIZE = 1000
CLASSES = utils.read_coco_names('./data/voc.names')
ANCHORS = utils.get_anchors('./data/voc_anchors.txt')
NUM_CLASSES = len(CLASSES)
train_tfrecord = "../VOC/train/voc_train*.tfrecords"
test_tfrecord = "../VOC/test/voc_test*.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
testset = dataset(parser, test_tfrecord, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = tf.cond(is_training, lambda: trainset.get_next(),
lambda: testset.get_next())
images, *y_true = example
import sys
import numpy as np
import tensorflow as tf
from tqdm import tqdm
from PIL import Image
from core import utils, yolov3
from core.dataset import dataset, Parser
import os
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = str(1)
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
CLASSES = utils.read_coco_names('./sixray.names') # todo
NUM_CLASSES = len(CLASSES)
ANCHORS = utils.get_anchors('./sixray_anchors.txt', IMAGE_H, IMAGE_W) # todo
CKPT_FILE = "./checkpoint/yolov3_sixray.ckpt-2500" # todo
IOU_THRESH = 0.5
SCORE_THRESH = 0.3
all_detections = []
all_annotations = []
all_aver_precs = {CLASSES[i]: 0. for i in range(NUM_CLASSES)}
test_tfrecord = "./sixray_*.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
testset = dataset(parser,
test_tfrecord,
batch_size=1,
import cv2
import numpy as np
import tensorflow as tf
from core import utils
from PIL import Image
from core.dataset import Parser, dataset
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 1
SHUFFLE_SIZE = 1
train_tfrecord = "./apple_dataset/apple_*.tfrecords"
anchors = utils.get_anchors('./data/apple_anchors.txt', IMAGE_H, IMAGE_W)
classes = utils.read_coco_names('./data/apple.names')
num_classes = len(classes)
parser = Parser(IMAGE_H, IMAGE_W, anchors, num_classes, debug=True)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
is_training = tf.placeholder(tf.bool)
example = trainset.get_next()
for l in range(20):
image, boxes = sess.run(example)
image, boxes = image[0], boxes[0]
n_box = len(boxes)
for i in range(n_box):
image = cv2.rectangle(image,(int(float(boxes[i][0])),
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
sess = tf.Session(config=config)
IMAGE_H, IMAGE_W = 416, 416 #输入网络模型图像尺寸
BATCH_SIZE = 8 # batch size 每次读取多少张图像
STEPS = 315200
#训练迭代的最大次数
#STEPS = 335000 #训练迭代的最大次数
LR = 0.0001 # if Nan, set 0.0005, 0.0001 初始学习率设置为0.0001
DECAY_STEPS = 100 #训练多少步之后对学习率进行一次调整
DECAY_RATE = 0.999 #学习率调整的比例
SHUFFLE_SIZE = 1000 #打散的个数
CLASSES = utils.read_coco_names('./data/mask.names') #类别数据
ANCHORS = utils.get_anchors('./data/mask_anchors.txt', IMAGE_H,
IMAGE_W) #归一化锚框
NUM_CLASSES = len(CLASSES) #类别数
EVAL_INTERNAL = 1000 #每100步评估一次 recall和precision
SAVE_INTERNAL = 1000 #每500步保存一次模型
train_tfrecord = "./train_data_tfrecord/train.tfrecords"
test_tfrecord = "./train_data_tfrecord/test.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
testset = dataset(parser, test_tfrecord, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = tf.cond(is_training, lambda: trainset.get_next(),
# Author : YunYang1994
# Created date: 2019-01-23 10:21:50
# Description :
#
#================================================================
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
import os
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = str(0)
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names('./sixray.names') # todo
num_classes = len(classes)
image_path = "/media/kent/DISK2/SBRI_Project/dataset_sixray/SIXray-master/ks3util-1.1.1-upload/JPEGImage/P08852.jpg" # 181,
img = Image.open(image_path)
img_resized = np.array(img.resize(size=(IMAGE_W, IMAGE_H)), dtype=np.float32)
img_resized = img_resized / 255.
cpu_nms_graph = tf.Graph()
input_tensor, output_tensors = utils.read_pb_return_tensors(
cpu_nms_graph, "./checkpoint/yolov3_cpu_nms.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
with tf.Session(graph=cpu_nms_graph) as sess:
boxes, scores = sess.run(
output_tensors,
feed_dict={input_tensor: np.expand_dims(img_resized, axis=0)})
boxes, scores, labels = utils.cpu_nms(boxes,
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "0"
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names('./data/raccoon.names')
num_classes = len(classes)
image_path = "data/hide_specific_target_dataset/000000336584.jpg" # 181,
img = Image.open(image_path)
img_resized = np.array(img.resize(size=(IMAGE_W, IMAGE_H)), dtype=np.float32)
img_resized = img_resized / 255.
cpu_nms_graph = tf.Graph()
input_tensor, output_tensors = utils.read_pb_return_tensors(
cpu_nms_graph, "./checkpoint/yolov3_cpu_nms.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
with tf.Session(graph=cpu_nms_graph) as sess:
boxes, scores = sess.run(
output_tensors,
feed_dict={input_tensor: np.expand_dims(img_resized, axis=0)})
boxes, scores, labels = utils.cpu_nms(boxes,
scores,
num_classes,
score_thresh=0.3,
iou_thresh=0.5)
p1x = box[0] * xDist
p1y = box[1] * yDist
p2x = box[2] * xDist
p2y = box[3] * yDist
return int(p1x), int(p1y), int(p2x), int(p2y)
classesPath = "/home/benoit/Documents/Stage2A/tensorflow-yolov3/data/coco.names"
modelPath = "/home/benoit/Documents/Stage2A/tensorflow-yolov3/checkpoint/yolov3_cpu_nms.pb"
destPath = "/home/benoit/Documents/Stage2A/resources/results/detection"
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names(classesPath)
num_classes = len(classes)
input_tensor, output_tensors = utils.read_pb_return_tensors(
tf.get_default_graph(), modelPath,
["Placeholder:0", "concat_9:0", "mul_6:0"])
videosPaths = {
"CP":
"/home/benoit/Documents/Stage2A/resources/CP_dataset/data/P2L_S5_C3.1/P2L_S5_C3.1.mp4",
# "PCDS": "/home/benoit/Documents/Stage2A/resources/PCDS_dataset/25_20160407_back/normal/crowd/2016_04_07_18_24_54BackColor.avi",
# "MIVIA": "/home/benoit/Documents/Stage2A/resources/MIVIA_dataset/Dataset People Counting MIVIA/DBc/VIDEOS/RGB/C_I_S_1.mkv",
# "MOT": "/home/benoit/Documents/Stage2A/resources/MOT_dataset/2DMOT2015/train/PETS09-S2L1/img1/PETS09-S2L1.mp4"
}
with tf.Session() as sess:
for name, videoPath in videosPaths.items():
import cv2
import numpy as np
import tensorflow as tf
from core import utils
from PIL import Image
from core.dataset import Parser, dataset
sess = tf.Session()
BATCH_SIZE = 1
SHUFFLE_SIZE = 1
CLASSES = utils.read_coco_names('./data/class.names')
ANCHORS = utils.get_anchors('./data/anchors.txt')
NUM_CLASSES = len(CLASSES)
TRAIN_TFRECORD = "./data/train_data/val.tfrecords"
TEST_TFRECORD = "./data/val_data/val.tfrecords"
parser_train = Parser(ANCHORS, NUM_CLASSES)
parser_test = Parser(ANCHORS, NUM_CLASSES)
trainset = dataset(parser_train, TRAIN_TFRECORD, BATCH_SIZE, shuffle=SHUFFLE_SIZE, multi_image_size=False)
testset = dataset(parser_test, TEST_TFRECORD, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = testset.get_next()
for l in range(1):
res = sess.run(example)
image = res[0][0] * 255
y_true = res[1:]
boxes = utils.decode_gtbox(y_true)
n_box = len(boxes)
for i in range(n_box):
image = cv2.rectangle(image,(int(float(boxes[i][0])),
int(float(boxes[i][1]))),
import cv2
import numpy as np
import tensorflow as tf
from core import utils
from PIL import Image
from core.dataset import Parser, dataset
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 1
SHUFFLE_SIZE = 1
train_tfrecord = "../data/voc_data/images_train.tfrecords"
anchors = utils.get_anchors('../data/objects.txt', IMAGE_H, IMAGE_W)
classes = utils.read_coco_names('../data/objects.names')
# print(classes)
num_classes = len(classes) # 识别的种类
parser = Parser(IMAGE_H, IMAGE_W, anchors, num_classes, debug=True)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
is_training = tf.placeholder(tf.bool)
example = trainset.get_next()
for l in range(1):
image, boxes = sess.run(example)
# print(image)
# print(sess.run(example))
image, boxes = image[0], boxes[0]
#! /usr/bin/env python
# coding=utf-8
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names('data/SVHN/SVHN.names')
num_classes = len(classes)
image_path = "data/SVHN/PaddingTest/1.png" # 181,
img = Image.open(image_path)
img_resized = np.array(img.resize(size=(IMAGE_W, IMAGE_H)), dtype=np.float32)
img_resized = img_resized / 255.
cpu_nms_graph = tf.Graph()
input_tensor, output_tensors = utils.read_pb_return_tensors(
cpu_nms_graph, "checkpoint/yolov3_cpu_nms.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
with tf.Session(graph=cpu_nms_graph) as sess:
boxes, scores = sess.run(
output_tensors,
feed_dict={input_tensor: np.expand_dims(img_resized, axis=0)})
boxes, scores, labels = utils.cpu_nms(boxes,
scores,
num_classes,
score_thresh=0.3,
iou_thresh=0.5)
image = utils.draw_boxes(img,
boxes,
scores,
# Editor : VIM
# File name : f**k.py
# Author : YunYang1994
# Created date: 2019-01-23 10:21:50
# Description :
#
#================================================================
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names('bdd100k/bdd100k.names')
num_classes = len(classes)
image_path = "D:/self-driving/tensorflow-yolov3-master/bdd100k/data/val/b4b99a3e-21dfc344.jpg" # 181,
img = Image.open(image_path)
img_resized = np.array(img.resize(size=(IMAGE_W, IMAGE_H)), dtype=np.float32)
img_resized = img_resized / 255.
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
cpu_nms_graph = tf.Graph()
input_tensor, output_tensors = utils.read_pb_return_tensors(cpu_nms_graph, "./checkpoint/yolov3_cpu_nms_bdd.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
with tf.Session(config=config,graph=cpu_nms_graph) as sess:
#
# Editor : VIM
# File name : f**k.py
# Author : YunYang1994
# Created date: 2019-01-23 10:21:50
# Description :
#
#================================================================
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
IMAGE_H, IMAGE_W = 416, 416
classes = utils.read_coco_names('./constructionsite_dataset/constructionsite.names')
num_classes = len(classes)
image_path = "./constructionsite_dataset/VOC2007/train/JPEGImages/23.jpg"
img = Image.open(image_path)
img_resized = np.array(img.resize(size=(IMAGE_W, IMAGE_H)), dtype=np.float32)
img_resized = img_resized / 255.
cpu_nms_graph = tf.Graph()
input_tensor, output_tensors = utils.read_pb_return_tensors(cpu_nms_graph, "./checkpoint/yolov3_cpu_nms.pb",
["Placeholder:0", "concat_9:0", "mul_6:0"])
with tf.Session(graph=cpu_nms_graph) as sess:
boxes, scores = sess.run(output_tensors, feed_dict={input_tensor: np.expand_dims(img_resized, axis=0)})
boxes, scores, labels = utils.cpu_nms(boxes, scores, num_classes, score_thresh=0.3, iou_thresh=0.5)
image = utils.draw_boxes(img, boxes, scores, labels, classes, [IMAGE_H, IMAGE_W], show=True)
import cfg
import numpy as np
import tensorflow as tf
from PIL import Image
from core import utils
from core.detect_function import draw_boxes, detection_detail
from core.yolov3 import YOLO_V3
classes = utils.read_coco_names(cfg.classes_names_file)
# with tf.variable_scope('yolov3'):
model = YOLO_V3()
sess = tf.Session()
for i, k in enumerate(tf.global_variables("yolov3"), start=1):
print(i, k.name)
sess.run(tf.global_variables_initializer())
# Loading pre_trained weights
print("Loading weights ...")
sess.run(utils.load_weights(tf.global_variables(scope='yolov3'), cfg.weights_file))
while True:
print("=============================================")
flag = 0
while flag == 0:
img_path = input("input the image path:")
try:
img = Image.open(img_path) # input RGB format
import numpy as np
import tensorflow as tf
from core import utils
from PIL import Image
from core.dataset import Parser, dataset
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 1
SHUFFLE_SIZE = 1
#train_tfrecord = "./raccoon_dataset/raccoon_*.tfrecords"
train_tfrecord = "data/tfdata/voc_*.tfrecords"
anchors = utils.get_anchors('data/voc_anchors.txt', IMAGE_H, IMAGE_W)
classes = utils.read_coco_names('data/voc.names')
num_classes = len(classes)
parser = Parser(IMAGE_H, IMAGE_W, anchors, num_classes, debug=True)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
is_training = tf.placeholder(tf.bool)
example = trainset.get_next()
for l in range(10):
image, boxes = sess.run(example)
image, boxes = image[0], boxes[0]
n_box = len(boxes)
for i in range(n_box):
image = cv2.rectangle(
# Description :
#
#================================================================
import tensorflow as tf
from core import utils, yolov3
INPUT_SIZE = 416
BATCH_SIZE = 1
EPOCHS = 20
LR = 0.001
SHUFFLE_SIZE = 1
weights_path = "/home/yang/test/yolov3.weights"
sess = tf.Session()
classes = utils.read_coco_names('./data/coco.names')
num_classes = len(classes)
file_pattern = "./data/train_data/quick_train_data/tfrecords/quick_train_data*.tfrecords"
anchors = utils.get_anchors('./data/yolo_anchors.txt')
dataset = tf.data.TFRecordDataset(filenames=tf.gfile.Glob(file_pattern))
dataset = dataset.map(utils.parser(anchors, num_classes).parser_example,
num_parallel_calls=10)
dataset = dataset.repeat().shuffle(SHUFFLE_SIZE).batch(BATCH_SIZE).prefetch(
BATCH_SIZE)
iterator = dataset.make_one_shot_iterator()
example = iterator.get_next()
images, *y_true = example
model = yolov3.yolov3(num_classes)
with tf.variable_scope('yolov3'):
# coding=utf-8
#================================================================
# Copyright (C) 2019 * Ltd. All rights reserved.
#
# Editor : Vscode
# File name : Config.py
# Author :
# Created date:
# Description :
#
#================================================================
from core import utils
IMAGE_H = 416
IMAGE_W = 416
CLASSES = utils.read_coco_names('./data/wider/wider.names')
NUM_CLASSES = len(CLASSES)
ANCHORS = utils.get_anchors('./data/wider/wider_anchors.txt', IMAGE_H, IMAGE_W)
#Training
BATCH_SIZE = 16
STEPS = 40000
Learning_Rate = 0.002 # beginning laerining
DECAY_STEPS = 5000 #
DECAY_RATE = 0.9 # every decay_steps,new_lr=current_lr *dacay_rate
SHUFFLE_SIZE = 200 # for data shuffle
MAX_TO_KEEP = 2 # training checkpoint
SAVE_INTERNAL = 400 # every each steps,save weights to checkpoint
EVAL_INTERNAL = 100 # every each steps, evaluate training result
Train_tfrecord = "./data/wider/train.tfrecords"
Valid_tfrecord = './data/wider/valid.tfrecords'
# Created date: 2018-12-20 11:58:21
# Description : compute mAP
#
#================================================================
import sys
import numpy as np
import tensorflow as tf
from tqdm import tqdm
from PIL import Image
from core import utils, yolov3
from core.dataset import dataset, Parser
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
CLASSES = utils.read_coco_names(
'./constructionsite_dataset/constructionsite.names')
NUM_CLASSES = len(CLASSES)
ANCHORS = utils.get_anchors(
'./constructionsite_dataset/constructionsite_anchors.txt', IMAGE_H,
IMAGE_W)
CKPT_FILE = "./checkpoint/yolov3.ckpt-2500"
IOU_THRESH = 0.5
SCORE_THRESH = 0.3
all_detections = []
all_annotations = []
all_aver_precs = {CLASSES[i]: 0. for i in range(NUM_CLASSES)}
test_tfrecord = "./tfrecords/voc_*.tfrecords"
# test_tfrecord = "./raccoon_dataset/raccoon_*.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
import tensorflow as tf
from core import utils, yolov3
import cv2
from PIL import Image
import numpy as np
input_image = "../data/raccoon_data/images/raccoon-4.jpg"
image = Image.open(input_image)
# image = cv2.imread(input_image)
# image = Image.fromarray(image)
image_resize = cv2.resize(np.array(image) / 255., (416, 416))
image_place = tf.placeholder(dtype=tf.float32, shape=(None, 416, 416, 3))
CLASSES = utils.read_coco_names('../data/objects.names')
ANCHORE = utils.get_anchors("../data/objects.txt", 416, 416)
model = yolov3.yolov3(len(CLASSES), ANCHORE)
with tf.variable_scope('yolov3'):
pred_feature_map = model.forward(image_place, is_training=False)
pred = model.predict(pred_feature_map)
sess = tf.Session()
saver = tf.train.Saver()
model_dir = tf.train.latest_checkpoint("../data/raccoon_data/model/")
saver.restore(sess, model_dir)
boxes, confs, prods = sess.run(pred, feed_dict={image_place: np.expand_dims(image_resize, 0)})
boxes, confs, prods = utils.cpu_nms(boxes, confs * prods, len(CLASSES))
utils.draw_boxes(image, boxes, confs, prods, CLASSES, (416, 416), "../data/font/HuaWenXinWei-1.ttf")
print(boxes, confs, prods)
#! /usr/bin/env python
# coding=utf-8
import sys
import numpy as np
import tensorflow as tf
from tqdm import tqdm
from PIL import Image
from core import utils, yolov3
from core.dataset import dataset, Parser
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
# CLASSES = utils.read_coco_names('./data/raccoon.names')
CLASSES = utils.read_coco_names('./data/apple.names')
NUM_CLASSES = len(CLASSES)
# ANCHORS = utils.get_anchors('./data/raccoon_anchors.txt', IMAGE_H, IMAGE_W)
ANCHORS = utils.get_anchors('./data/apple_anchors.txt', IMAGE_H, IMAGE_W)
CKPT_FILE = "./checkpoint2/yolov3.ckpt-10000"
IOU_THRESH = 0.5
SCORE_THRESH = 0.3
all_detections = []
all_annotations = []
all_aver_precs = {CLASSES[i]:0. for i in range(NUM_CLASSES)}
# test_tfrecord = "./raccoon_dataset/raccoon_*.tfrecords"
# test_tfrecord = "./apple_dataset/apple_*.tfrecords"
test_tfrecord = "./apple_dataset/apple_test.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
import tensorflow as tf
from core import utils, yolov3
from core.dataset import dataset, Parser
#此部分为迁移学习进行训练
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 8
STEPS = 2500
LR = 0.001 # if Nan, set 0.0005, 0.0001
DECAY_STEPS = 100
DECAY_RATE = 0.9
SHUFFLE_SIZE = 200
CLASSES = utils.read_coco_names('../data/raccoon.names')
ANCHORS = utils.get_anchors('../data/raccoon_anchors.txt', IMAGE_H, IMAGE_W)
NUM_CLASSES = len(CLASSES)
EVAL_INTERNAL = 100
SAVE_INTERNAL = 500
train_tfrecord = "../data/train_dome_data/images_train.tfrecords"
test_tfrecord = "../data/train_dome_data/images_test.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
testset = dataset(parser, test_tfrecord, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool) #占位符 方便session传递参数
example = tf.cond(is_training, lambda: trainset.get_next(),
lambda: testset.get_next())
# y_true = [feature_map_1 , feature_map_2 , feature_map_3]
from PIL import Image
from core import utils, yolov3
from core.dataset import dataset, Parser
coco_ids = [
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22,
23, 24, 25, 27, 28, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44,
46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64,
65, 67, 70, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 84, 85, 86, 87, 88,
89, 90
]
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
CLASSES = utils.read_coco_names('./data/coco.names')
NUM_CLASSES = len(CLASSES)
ANCHORS = utils.get_anchors('./data/coco_anchors.txt', IMAGE_H, IMAGE_W)
CKPT_FILE = "/home/common/pretrained_models/checkpoint/yolov3.ckpt"
IOU_THRESH = 0.5
SCORE_THRESH = 0.001
all_detections = []
all_annotations = []
all_aver_precs = {CLASSES[i]: 0. for i in range(NUM_CLASSES)}
test_tfrecord = "/home/common/datasets/tfrecords/5k.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
testset = dataset(parser,
test_tfrecord,
batch_size=1,
def main():
global config
OutDir = './data/HideSpecificTarget_result/1/'
Img_path = "data/coco_test_data/2.jpg"
attack_index = 0 # 攻击第几个框 from 0
continuous_round = 20
mark = 0
if os.path.exists(OutDir):
shutil.rmtree(OutDir)
os.makedirs(OutDir)
pNumber = 3
MaxEpoch = 20000 # 迭代上限
Phi1 = 0.4
Phi2 = 1.2
Phi3 = 4
# Phi4 = 1/20
# Phi5 = 1/70
IMAGE_H = 416
IMAGE_W = 416
with tf.Session(config=config) as sess:
classes = utils.read_coco_names('./data/coco.names')
###########对原图进行剪切获取SImg###############
yolo = YoloTest()
source_img = Image.open(Img_path)
TImg = np.array(source_img.resize((IMAGE_H, IMAGE_W)), dtype=np.float32)/255.
yolo.predict_from_array(TImg)
yolo.last_PIL_image = Image.fromarray((255. * TImg).astype('uint8')).convert('RGB')
yolo.save_result(os.path.join(OutDir, 'yuanshitupian.jpg' ))
original_boxes, scores, labels = yolo.predict_from_array(TImg)
assert len(labels) > attack_index
# attack first label
#attack_index = random.randint(0, len(labels) - 1)
attack_index = random.choice(range(len(labels)))
print("attack random label: {} ".format(classes[labels[attack_index]]))
attcked_label = labels[attack_index]
x_min = original_boxes[attack_index][0]
y_min = original_boxes[attack_index][1]
x_max = original_boxes[attack_index][2]
y_max = original_boxes[attack_index][3]
center_point = ((x_min+x_max)/2, (y_max+y_min)/2)
Top_One_SM = np.zeros(shape=(pNumber), dtype=float)
pbestFitness = np.zeros(shape=(pNumber), dtype=float)
gbest = np.zeros(shape=(IMAGE_H, IMAGE_W, 3), dtype=float)
gbestFitness = -1e8
xOldOld = np.repeat(np.expand_dims(TImg, axis=0), pNumber, axis=0)
xOld = np.repeat(np.expand_dims(TImg, axis=0), pNumber, axis=0)
pbest = np.zeros(shape=(pNumber, IMAGE_H, IMAGE_W, 3), dtype=float)
x = np.zeros(shape=(pNumber, IMAGE_H, IMAGE_W, 3), dtype=float)
L2D = np.zeros(shape=(pNumber), dtype=float)
Top_One_SM_pbest = np.zeros(shape=(pNumber), dtype=float)
L2D_pbest = np.zeros(shape=(pNumber), dtype=float)
Top_One_SM_gbest = 0.0
L2D_gbest = 0.0
sign_gbest = 0 # 标记是否有新的gbest生成
Top_One_SM_SUM_round = np.ones(shape=(continuous_round), dtype=float) # 记录每一轮的Top_One_SM_SUM
x_round = np.zeros(shape=(continuous_round, pNumber, IMAGE_H, IMAGE_W, 3), dtype=float) # 记录每一轮的x
xOld_round = np.zeros(shape=(continuous_round, pNumber, IMAGE_H, IMAGE_W, 3), dtype=float) # 记录每一轮的xOld
xOldOld_round = np.zeros(shape=(continuous_round, pNumber, IMAGE_H, IMAGE_W, 3), dtype=float) # 记录每一轮的xOldOld
gbest_round = np.zeros(shape=(continuous_round, pNumber, IMAGE_H, IMAGE_W, 3), dtype=float) # 记录每一轮的gbest
for i in range(MaxEpoch):
print("[Epoch {}]################################################################################################################".format(i))
if i==0:
# initialization
print( "initialization")
for l in range(pNumber):
x[l] = TImg + 2 * np.random.randn(IMAGE_H, IMAGE_W, 3)
x[l] = np.clip(x[l], 0.0, 1.0)
L2D[l] = np.linalg.norm(TImg - x[l])
xOld = np.copy(x)
xOldOld = np.copy(xOld)
x_round[i] = np.copy(x)
xOld_round[i] = np.copy(xOld)
xOldOld_round[i] = np.copy(xOldOld)
##########计算 标签得分##############
for j in range(pNumber):
boxes, scores, labels = yolo.predict_from_array(x[j])
Top_One_SM[j] = 50
for m, label in enumerate(labels):
if label == attcked_label:
# 如果原框还在，依据得分惩罚它
if is_point_in_box(center_point, boxes[m]):
Top_One_SM[j] = 0
break
Top_One_SM_SUM = 0
for b in range(pNumber):
Top_One_SM_SUM = Top_One_SM_SUM + Top_One_SM[b]
Top_One_SM_SUM_round[i] = Top_One_SM_SUM
fitness = Top_One_SM - L2D
pbestFitness = np.copy(fitness)
pbest = np.copy(x)
Top_One_SM_pbest = np.copy(Top_One_SM)
L2D_pbest = np.copy(L2D)
for m in range(pNumber):
if pbestFitness[m] > gbestFitness:
gbestFitness = pbestFitness[m]
gbest = np.copy(pbest[m])
Top_One_SM_gbest = Top_One_SM[m]
L2D_gbest = L2D[m]
gbest_round[i] = np.copy(gbest)
yolo.predict_from_array(gbest)
yolo.last_PIL_image = Image.fromarray((255. * gbest).astype('uint8')).convert('RGB')
yolo.save_result(os.path.join(OutDir, 'aaaaagbest_query_%d_TopOneSm_%.4f_L2_%.1f.jpg' %(i, Top_One_SM_gbest, L2D_gbest)))
print("l2distance:", L2D )
print("Label Score", Top_One_SM)
else:
if mark == 0:
Top_One_SM_SUM_round_Sum = 0
# 求 Top_One_SM_SUM_round中第1到最后一个元素的和
for d in range(1, continuous_round):
Top_One_SM_SUM_round_Sum += Top_One_SM_SUM_round[d]
if bool(1 - (Top_One_SM_SUM_round[0] != 0 and Top_One_SM_SUM_round_Sum == 0)):
pv = Phi1 * ((xOld - xOldOld) / 2 + Phi2 * np.random.rand() * (pbest - x) + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / 20 + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 70)
x = x + pv
x = np.clip(x, 0.0, 1.0)
print('Now in the Rapid iteration###################################################################################')
if i <= continuous_round - 1:
x_round[i] = np.copy(x)
else:
x_round[:continuous_round - 1] = np.copy(x_round[1:])
x_round[continuous_round - 1] = np.copy(x)
else:
x = x_round[0]
xOld = xOld_round[0]
xOldOld = xOldOld_round[0]
gbest = gbest_round[0]
mark = 1
else:
if i < 30:
guanxing = Phi1 * (xOld - xOldOld) / 2
pbest_x = Phi1 * Phi2 * np.random.rand() * (pbest - x)
gbest_x = Phi1 * Phi3 * np.random.rand() * (gbest - x)
SImg_x = Phi1 * (TImg - x) / 100
rand_num_list = Phi1 * np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 100
pv = guanxing + pbest_x + gbest_x + SImg_x + rand_num_list
x = x + pv
x = np.clip(x, 0.0, 1.0)
print('x:', x[0][0][0])
print('guanxing:', guanxing[0][0][0])
print('pbest_x:', pbest_x[0][0][0])
print('gbest_x:', gbest_x[0][0][0])
print('SImg_x:', SImg_x[0][0][0])
print('rand_num_list:', rand_num_list[0][0][0])
print('pv:', pv[0][0][0])
elif i < 100:
pv = Phi1 * ((xOld - xOldOld) / 2 + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / 200 + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 70)
x = x + pv
x = np.clip(x, 0.0, 1.0)
elif i < 600:
pv = Phi1 * ((xOld - xOldOld) / 4 + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / (0.65 * i + 170) + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 70)
x = x + pv
x = np.clip(x, 0.0, 1.0)
elif i < 1000:
pv = Phi1 * ((xOld - xOldOld) / 5 + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / 400 + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 100) # 此处pv并没有加约束，v使用的是标准正态分布，可以更改
x = x + pv
x = np.clip(x, 0.0, 1.0)
elif i < 1500:
pv = Phi1 * ((xOld - xOldOld) / 5 + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / 600 + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 200) # 此处pv并没有加约束，v使用的是标准正态分布，可以更改
x = x + pv
x = np.clip(x, 0.0, 1.0)
else:
pv = Phi1 * ((xOld - xOldOld) / 5 + Phi3 * np.random.rand() * (gbest - x) + (TImg - x) / 800 + np.random.randn(pNumber, IMAGE_H, IMAGE_W, 3) / 300) # 此处pv并没有加约束，v使用的是标准正态分布，可以更改
x = x + pv
x = np.clip(x, 0.0, 1.0)
xOldOld = np.copy(xOld)
xOld = np.copy(x)
for j in range(pNumber):
L2D[j] = np.linalg.norm(TImg - x[j])
boxes, scores, labels = yolo.predict_from_array(x[j])
Top_One_SM[j] = 50
for m, label in enumerate(labels):
if label == attcked_label:
# 如果原框还在，依据得分惩罚它
if is_point_in_box(center_point, boxes[m]):
Top_One_SM[j] = 0
break
fitness = Top_One_SM - L2D
if mark == 0:
Top_One_SM_SUM = 0
for b in range(pNumber):
Top_One_SM_SUM = Top_One_SM_SUM + Top_One_SM[b]
if i <= continuous_round - 1:
xOld_round[i] = np.copy(xOld)
xOldOld_round[i] = np.copy(xOldOld)
Top_One_SM_SUM_round[i] = Top_One_SM_SUM
else:
xOld_round[:continuous_round - 1] = np.copy(xOld_round[1:])
xOld_round[continuous_round - 1] = np.copy(xOld)
xOldOld_round[:continuous_round - 1] = np.copy(xOldOld_round[1:])
xOldOld_round[continuous_round - 1] = np.copy(xOldOld)
Top_One_SM_SUM_round[:continuous_round - 1] = Top_One_SM_SUM_round[1:]
Top_One_SM_SUM_round[continuous_round - 1] = Top_One_SM_SUM
for e in range(pNumber):
##########更新pbest和pbestFitness###############
if fitness[e] > pbestFitness[e]:
pbestFitness[e] = fitness[e]
pbest[e] = np.copy(x[e])
Top_One_SM_pbest[e] = Top_One_SM[e]
L2D_pbest[e] = L2D[e]
for e in range(pNumber):
##########更新gbest和gbestFitness###############
if pbestFitness[e] > gbestFitness:
sign_gbest = 1
gbestFitness = pbestFitness[e]
gbest = np.copy(pbest[e])
Top_One_SM_gbest = Top_One_SM_pbest[e]
L2D_gbest = L2D_pbest[e]
if mark == 0:
if i <= continuous_round - 1:
gbest_round[i] = np.copy(gbest)
else:
gbest_round[:continuous_round - 1] = np.copy(gbest_round[1:])
gbest_round[continuous_round - 1] = np.copy(gbest)
if sign_gbest == 1:
yolo.predict_from_array(gbest)
yolo.last_PIL_image = Image.fromarray((255. * gbest).astype('uint8')).convert('RGB')
yolo.save_result(os.path.join(OutDir, 'aaaaagbest_query_%d_TopOneSm_%.4f_L2_%.1f.jpg' % (i, Top_One_SM_gbest, L2D_gbest)))
sign_gbest = 0
# print()
print("l2distance:", L2D)
print("Label Score:", Top_One_SM)
#! /usr/bin/env python
# coding=utf-8
import tensorflow as tf
from core import utils, yolov3
from core.dataset import dataset, Parser
from PIL import Image
sess = tf.Session()
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 1
SHUFFLE_SIZE = 200
CLASSES = utils.read_coco_names("data/SVHN/SVHN.names")
ANCHORS = utils.get_anchors("data/SVHN/SVHN_anchors.txt", IMAGE_H, IMAGE_W)
NUM_CLASSES = len(CLASSES)
test_tfrecord = "data/SVHN/tfrecords/quick_test_data.tfrecords"
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
testset = dataset(parser, test_tfrecord, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = testset.get_next()
images, *y_true = example
model = yolov3.yolov3(NUM_CLASSES, ANCHORS)
with tf.variable_scope('yolov3'):
pred_feature_map = model.forward(images, is_training=is_training)
loss = model.compute_loss(pred_feature_map, y_true)
y_pred = model.predict(pred_feature_map)
class_name_path = "bdd100k/bdd100k.names"
train_tfrecord = "bdd100k/train_tf.tfrecords"
test_tfrecord = "bdd100k/val_tf.tfrecords"
dataname="bdd100k"
IMAGE_H, IMAGE_W = 416, 416
BATCH_SIZE = 8
STEPS = 25000
LR = 0.001 # if Nan, set 0.0005, 0.0001
DECAY_STEPS = 100
DECAY_RATE = 0.9
SHUFFLE_SIZE = 200
CLASSES = utils.read_coco_names(class_name_path)
ANCHORS = utils.get_anchors('data/voc_anchors.txt', IMAGE_H, IMAGE_W)
NUM_CLASSES = len(CLASSES)
EVAL_INTERNAL = 100
SAVE_INTERNAL = 500
parser = Parser(IMAGE_H, IMAGE_W, ANCHORS, NUM_CLASSES)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
testset = dataset(parser, test_tfrecord , BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = tf.cond(is_training, lambda: trainset.get_next(), lambda: testset.get_next())
images, *y_true = example
import cv2
import numpy as np
import tensorflow as tf
from core import utils
from PIL import Image
from core.dataset import Parser, dataset
sess = tf.Session()
INPUT_SIZE = 416
BATCH_SIZE = 1
SHUFFLE_SIZE = 1
train_tfrecord = "/home/yang/test/highwaydata/tfrecords/highwaydata*.tfrecords"
test_tfrecord = "/home/yang/test/highwaydata/tfrecords/highwaydata*.tfrecords"
anchors = utils.get_anchors('./data/coco_anchors.txt')
classes = utils.read_coco_names('./data/vehicle.names')
num_classes = len(classes)
parser = Parser(416, 416, anchors, num_classes, debug=True)
trainset = dataset(parser, train_tfrecord, BATCH_SIZE, shuffle=SHUFFLE_SIZE)
testset = dataset(parser, test_tfrecord, BATCH_SIZE, shuffle=None)
is_training = tf.placeholder(tf.bool)
example = tf.cond(is_training, lambda: trainset.get_next(),
lambda: testset.get_next())
for l in range(20):
image, boxes = sess.run(example, feed_dict={is_training: False})
image, boxes = image[0], boxes[0]
n_box = len(boxes)
