|
- # -*- coding:utf-8 -*-
-
- import tensorflow as tf
- import cv2
- from tqdm import *
- import time
- from utils.data_utils import create_iterator
-
-
- def test_dim_size():
- # slim = tf.contrib.slim
- input = tf.Variable(tf.random_uniform([1, 5, 5, 3]))
- kernel1 = tf.Variable(tf.random_uniform([1, 1, 3, 1])) # 1,1,3,1
- kernel2 = tf.concat([kernel1, kernel1], 3) # 1,1,3,3
- out1 = tf.nn.conv2d(input, kernel1, strides=[1, 1, 1, 1], padding='VALID')
- out2 = tf.nn.conv2d(input, kernel2, strides=[1, 1, 1, 1], padding='VALID')
-
- with tf.Session() as sess:
- tf.global_variables_initializer().run()
- print("\ninput----->\n", input.eval())
- print("\nkernel----->\n", kernel1.eval())
- print("\nkernel2----->\n", kernel2.eval())
- print("\n----->\n", out1.eval())
- print("\n----->\n", out2.eval())
-
-
- def test_dataset():
- train_init_op, val_init_op, image_ids, image, y_true = create_iterator()
- with tf.Session() as sess:
- sess.run(train_init_op)
- for i in range(2):
- sess.run(image)
-
-
- def test_txt_write():
- first = []
- second = []
- f = open('mergeTXT.txt', 'w')
- with open('first.txt', 'r') as f1:
- for line in f1:
- line = line.strip()
- first.append(line)
- with open('second.txt', 'r') as f2:
- for line2 in f2:
- line2 = line2.strip()
- second.append(line2)
- for i in range(0, 399):
- result = first[i] + '\t' + second[i] + '\n'
- f.write(result)
-
-
- def test_tqdm():
- with tqdm(total=100) as pbar:
- for i in range(10):
- time.sleep(1)
- pbar.update(10)
-
-
- def test_tqdm2():
- with trange(10000) as t:
- for i in t:
- t.set_description('下载速度 %i' % i)
-
-
-
- def test_plot_bbox():
- img = cv2.imread('data/demo_data/dog.jpg')
- #
- cv2.rectangle(img, (10, 100), (20, 200), (0, 255, 0), 2)
- cv2.imshow('img_detect', img)
- cv2.waitKey(0)
-
-
- def test_watch_save_weights():
- from tensorflow.python import pywrap_tensorflow
- # model_dir = 'checkpoint/model-epoch_12_step_64_loss_1.9270_lr_0.0001'
- model_dir = 'data/darknet_weights/yolov3.ckpt'
- reader = pywrap_tensorflow.NewCheckpointReader(model_dir)
- var_to_shape_map = reader.get_variable_to_shape_map()
- print("have {} tensor".format(len(var_to_shape_map)))
- for key in var_to_shape_map:
- print("tensor_name:{}, shape:{}".format(key, reader.get_tensor(key).shape))
|