tensorflow學習025——常見的預訓練網路模型及使用示例
在ImageNet上與訓練過的用於影象分類的模型：VGG16, VGG19, ResNet50, InceptionV3, InceptionResNetV2, Xception, MobileNet, MobileNetV2, DenseNet, NASNet
是在ImageNet上1000個分類的準確率
top1——你預測的label取最後概率向量裡面最大的那一個作為預測結果，如果你的預測結果中概率最大的那個分類正確，則預測正確，否則預測錯誤
top5——就是最後概率向量最大的前五名中，只要出現了正確預測，則為預測準確，否則預測錯誤。
從上面途中我們可以看出，VGG16 VGG19是比較落後的，訓練引數多但是準確率低。MobileNet MobileNetV2 僅僅十幾M，但是精度也不低，可以部署於手機上。
在ImageNet上預訓練的XceptionV1模型，在ImageNet上，該模型取得了驗證集top1 0.790和top 5 0.945的準確率。需要注意的是該模型只支援channels_last的維度順序（高度、寬度、通道），該模型預設輸入尺寸是299*299
其它訓練網路引數可通過網址(
Xception網路訓練貓狗資料集程式碼
資料鏈接：https://pan.baidu.com/s/1-nLlW6Nng1pAvrxwEfHttA
提取碼：vt8p
點選檢視程式碼
import tensorflow as tf import matplotlib.pyplot as plt import numpy as np import glob import os train_image_path = glob.glob(r'E:\WORK\tensorflow\dataset\dc_2000\train\*\*.jpg') print(len(train_image_path)) print(train_image_path[:5]) train_image_label = [int(p.split("\\")[-2] == 'cat') for p in train_image_path] def load_preprosess_image(path,label): image = tf.io.read_file(path) image = tf.image.decode_jpeg(image,channels=3) image = tf.image.resize(image,[256,256]) image = tf.cast(image,tf.float32) image = image / 255 return image,label train_image_ds = tf.data.Dataset.from_tensor_slices((train_image_path,train_image_label)) AUTOTUNE = tf.data.experimental.AUTOTUNE train_image_ds = train_image_ds.map(load_preprosess_image,num_parallel_calls=AUTOTUNE) BATCH_SZIE = 32 train_count = len(train_image_path) train_image_ds = train_image_ds.shuffle(train_count).repeat().batch(BATCH_SZIE) test_image_path = glob.glob(r'E:\WORK\tensorflow\dataset\dc_2000\test\*\*.jpg') test_image_label = [int(p.split("\\")[-2] == 'cat') for p in test_image_path] test_image_ds = tf.data.Dataset.from_tensor_slices((test_image_path,test_image_label)) test_image_ds = test_image_ds.map(load_preprosess_image,num_parallel_calls=AUTOTUNE) test_image_ds = test_image_ds.repeat().batch(BATCH_SZIE) test_count = len(test_image_path) conv_base = tf.keras.applications.xception.Xception(weights='imagenet', include_top=False, input_shape=(256,256,3), pooling='avg') conv_base.trainable = False model = tf.keras.Sequential() model.add(conv_base) model.add(tf.keras.layers.Dense(512,activation='relu')) model.add(tf.keras.layers.Dense(1,activation='sigmoid')) model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.0005), loss='binary_crossentropy', metrics=['acc']) initial_epoches = 5 histroy = model.fit(train_image_ds, steps_per_epoch=train_count//BATCH_SZIE, epochs=initial_epoches, validation_data=test_image_ds, validation_steps=test_count//BATCH_SZIE) conv_base.trainable = True for layer in conv_base.layers[:-33]: # 原先弓133層 layer.trainable = False model.compile(loss='binary_crossentropy', optimizer=tf.keras.optimizers.Adam(learning_rate=0.0005/10), metrics=['accuracy']) fine_tune_epoches = 5 total_epoches = initial_epoches + fine_tune_epoches histroy = model.fit(train_image_ds, steps_per_epoch=train_count//BATCH_SZIE, epochs=total_epoches, initial_epoch=initial_epoches, validation_data=test_image_ds, validation_steps=test_count//BATCH_SZIE)
作者：孫建釗
出處：http://www.cnblogs.com/sunjianzhao/
本文版權歸作者和部落格園共有，歡迎轉載，但未經作者同意必須保留此段宣告，且在文章頁面明顯位置給出原文連線，否則保留追究法律責任的權利。