コード例 #1
0
img_root = '/media/ubuntu/65db2e03-ffde-4f3d-8f33-55d73836211a/dataset/VOCdevkit/VOC2007/Test/JPEGImages'
labelfiles = '/media/ubuntu/65db2e03-ffde-4f3d-8f33-55d73836211a/dataset/VOCdevkit/VOC2007/Test/ImageSets/Main'
checkpoint_dir = '../../model/yolol2sum_epoch_SGD'
classes = voc.list_image_sets()
val_list = voc.imgs_from_category_as_list('', 'test', labelfiles)

yolo_old = YOLO_tiny_tf.YOLO_TF()
with tf.device('/gpu:0'):
    #Vanilla YOLO_tiny Weight
    x = tf.placeholder(tf.float32, (None, 448, 448, 3))
    label = tf.placeholder(tf.float32, (None, 1470), name='labels')
    keep_prob = tf.placeholder(tf.float32)

    modelTicket_G = {'root': 'yolo_tiny', 'branch': 'vanilla'}
    init_layers = mu.model_zoo(modelTicket_G)
    var_dict = recursive_create_var('recursive', 1, 0.2, init_layers)
    yolo_ds = nf.glosso_train("recursive_0", 'test', x, var_dict, keep_prob,
                              False)

    tlossTicket = {'loss': 'smoothL1'}
    loss_pair = {'prob': yolo_ds}
    loss = mu.loss_zoo(tlossTicket, loss_pair, label)

tp = 0
fp = 0

tp_old = 0
fp_old = 0

num = 0
コード例 #2
0
import tensorflow as tf
import numpy as np
import yolo_netfactory as nf
import random_batch as rb
import YOLO_tiny_tf
import cv2
import model_utility as mut
import time

#Vanilla Yolo
scope = 'train'
yolo = YOLO_tiny_tf.YOLO_TF()

#Target Model
model_ticket = {'root': yolo_tiny, 'branch': 'double_cut89'}
ds_yolo = mut.create_var_tnorm(scope, mut.model_zoo(model_ticket))

batch_file = "/media/ubuntu/65db2e03-ffde-4f3d-8f33-55d73836211a/dataset/VOC_train"
test_file = "/media/ubuntu/65db2e03-ffde-4f3d-8f33-55d73836211a/dataset/VOC_val"
filename = "../../model/test/fcann_v1.ckpt"
logfile = '../../log/test'
graph_model = '../../model/test/fcann_v1.ckpt-4000.meta'
checkpoint_dir = '../../model/test'

continue_training = 1
loop_num = 5900
batch_size = 64

keep_prob = tf.placeholder(tf.float32)
x = tf.placeholder(tf.float32, (None, 448, 448, 3))
label = tf.placeholder(tf.float32, (None, 1470))  #
コード例 #3
0
save_epoch = 200
test_epoch = 500
weight_decay = 0.0005

yolo = YOLO_tiny_tf.YOLO_TF()

with tf.device('/gpu:0'):

    keep_prob = tf.placeholder(tf.float32, name='dropout_prob')
    x = tf.placeholder(tf.float32, (None, 448, 448, 3), name='input_batch')
    label = tf.placeholder(tf.float32, (None, 1470), name='labels')
    tlabel = tf.placeholder(tf.float32, (None, 1470), name='test_labels')
    learning_rate = tf.placeholder(tf.float32, shape=[], name='learning_rate')

    model_ticket = {'root': 'yolo_tiny', 'branch': 'vanilla'}
    init_layers = mu.model_zoo(model_ticket)
    var_dict = recursive_create_var('recursive', 1, 0.2, init_layers)
    var_list = var_dict['recursive_0']

    with tf.name_scope('Weight_sum'):
        with tf.variable_scope('recursive_0') as scope:
            scope.reuse_variables()
            weight_sum = tf.reduce_sum([
                0.5 *
                tf.reduce_sum(tf.square(tf.get_variable(x) * weight_decay))
                for x in var_list
            ])
            w1 = tf.get_variable(var_list[0])

    glosso_train = nf.glosso_train("recursive_0", 'train', x, var_dict,
                                   keep_prob, True)
コード例 #4
0
save_epoch = 200
test_epoch = 500

modelTicket_G = {'root':'yolo_tiny', 'branch':'double_cut89'}
modelTicket_D = {'root':'discriminator', 'branch':'4layer'}


keep_prob = tf.placeholder(tf.float32)
x = tf.placeholder(tf.float32,(None,448,448,3))
test = tf.placeholder(tf.float32,(None,448,448,3))
label = tf.placeholder(tf.float32,(None,1470))

yolo = YOLO_tiny_tf.YOLO_TF()


gen_var = mut.create_var_xavier('train',mut.model_zoo(modelTicket_G))
dis_var = mut.create_var_xavier('discriminator', mut.model_zoo(modelTicket_D))

theta_D = []
theta_G = []

for i in mut.model_zoo(modelTicket_G):
    theta_G.append(gen_var[i[0]])
for i in mut.model_zoo(modelTicket_D):
    theta_D.append(dis_var[i[0]])


##Train Phase
yolo_ds_train = nf.yolo_dinception("yolo_train", x, gen_var, keep_prob, True)
lossTicket = {'loss':'L2norm'}
loss = mut.loss_zoo(lossTicket, yolo_ds_train, label)