迁移学习-VGG19手势识别.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import h5py\n",
    "import numpy as np\n",
    "import keras\n",
    "from tensorflow.python.keras.utils.np_utils import to_categorical\n",
    "from keras.applications.vgg19 import VGG19\n",
    "from keras.models import Model\n",
    "from keras.layers import Dense,GlobalMaxPooling2D,Input\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----Reshaped data:\n",
      "(1080, 64, 64, 3)\n",
      "(1080, 6)\n",
      "(120, 64, 64, 3)\n",
      "(120, 6)\n"
     ]
    }
   ],
   "source": [
    "# 导入数据（https://nekokecore.lanzoub.com/iHg4L086j15a）\n",
    "train_dataset = h5py.File('datasets/train_signs.h5', \"r\")\n",
    "train_set_x_orig = np.array(train_dataset[\"train_set_x\"][:]) \n",
    "train_set_y_orig = np.array(train_dataset[\"train_set_y\"][:]) \n",
    "test_dataset = h5py.File('datasets/test_signs.h5', \"r\")\n",
    "test_set_x_orig = np.array(test_dataset[\"test_set_x\"][:]) \n",
    "test_set_y_orig = np.array(test_dataset[\"test_set_y\"][:]) \n",
    "\n",
    "classes = np.array(test_dataset[\"list_classes\"][:]) \n",
    "# the list of classes\n",
    "train_set_y_orig = train_set_y_orig.reshape((train_set_y_orig.shape[0],1))\n",
    "test_set_y_orig = test_set_y_orig.reshape((test_set_y_orig.shape[0],1))\n",
    "\n",
    "print(\"-----Reshaped data:\")\n",
    "X_train = train_set_x_orig/255\n",
    "X_test = test_set_x_orig/255\n",
    "Y_train = to_categorical(train_set_y_orig,6)\n",
    "Y_test = to_categorical(test_set_y_orig,6)\n",
    "print(X_train.shape)\n",
    "print(Y_train.shape)\n",
    "print(X_test.shape)\n",
    "print(Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x15eb4ffe5e0>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(train_set_x_orig[30])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_model = VGG19(weights='imagenet',include_top=False)\n",
    "\n",
    "x = base_model.output\n",
    "x = GlobalMaxPooling2D()(x)\n",
    "x = Dense(128,activation='relu')(x)\n",
    "predictions = Dense(6,activation='softmax')(x)\n",
    "model = Model(inputs=base_model.input,outputs=predictions)\n",
    "\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "17/17 [==============================] - 5s 65ms/step - loss: 1.4391 - accuracy: 0.4750\n",
      "Epoch 2/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.8986 - accuracy: 0.7546\n",
      "Epoch 3/20\n",
      "17/17 [==============================] - 0s 25ms/step - loss: 0.6940 - accuracy: 0.7796\n",
      "Epoch 4/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.5777 - accuracy: 0.8241\n",
      "Epoch 5/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.4927 - accuracy: 0.8546\n",
      "Epoch 6/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.4199 - accuracy: 0.9056\n",
      "Epoch 7/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.3633 - accuracy: 0.9241\n",
      "Epoch 8/20\n",
      "17/17 [==============================] - 0s 23ms/step - loss: 0.3245 - accuracy: 0.9333\n",
      "Epoch 9/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.2927 - accuracy: 0.9435\n",
      "Epoch 10/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.2686 - accuracy: 0.9509\n",
      "Epoch 11/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.2533 - accuracy: 0.9426\n",
      "Epoch 12/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.2181 - accuracy: 0.9639\n",
      "Epoch 13/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.2039 - accuracy: 0.9676\n",
      "Epoch 14/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1978 - accuracy: 0.9583\n",
      "Epoch 15/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1863 - accuracy: 0.9648\n",
      "Epoch 16/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1651 - accuracy: 0.9657\n",
      "Epoch 17/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1498 - accuracy: 0.9778\n",
      "Epoch 18/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1431 - accuracy: 0.9759\n",
      "Epoch 19/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1329 - accuracy: 0.9796\n",
      "Epoch 20/20\n",
      "17/17 [==============================] - 0s 24ms/step - loss: 0.1216 - accuracy: 0.9824\n",
      "4/4 [==============================] - 1s 138ms/step - loss: 0.2102 - accuracy: 0.9583\n",
      "Total loss: 0.2101897895336151\n",
      "Test accuracy: 0.9583333134651184\n"
     ]
    }
   ],
   "source": [
    "model.fit(X_train,Y_train,epochs=20,batch_size=64)\n",
    "score = model.evaluate(X_test,Y_test)\n",
    "print(\"Total loss:\",score[0])\n",
    "print(\"Test accuracy:\",score[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "手势:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "预测值： 2\n",
      "手势:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "预测值： 0\n"
     ]
    }
   ],
   "source": [
    "# 0~119\n",
    "def pre(index):\n",
    "    print(\"手势:\")\n",
    "    plt.imshow(X_test[index])\n",
    "    plt.show()\n",
    "    print(\"预测值：\",np.argmax(model.predict(X_test[index:index+1])))\n",
    "\n",
    "pre(100)\n",
    "pre(1)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "vscode": {
   "interpreter": {
    "hash": "f7e751bf8ac0e8fd03b88115207dac510109b5859a0c795a0c781e2cefb625fe"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}