pfaure/Projet_innovant_IA
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

Modifs archi données

Browse source
This commit is contained in:
Clémentine Bonneau 2021-12-22 20:45:07 +01:00
parent 4312087dc9
commit fbea931fba
13 changed files with 76 additions and 63 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

139
Oiseau.ipynb
View file

@ -46,7 +46,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": 1,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -81,9 +81,31 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": 5,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[name: \"/device:CPU:0\"\n",
"device_type: \"CPU\"\n",
"memory_limit: 268435456\n",
"locality {\n",
"}\n",
"incarnation: 6212331521511793857\n",
"]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"2021-12-22 20:34:12.391220: I tensorflow/core/platform/cpu_feature_guard.cc:145] This TensorFlow binary is optimized with Intel(R) MKL-DNN to use the following CPU instructions in performance critical operations: SSE4.1 SSE4.2\n",
"To enable them in non-MKL-DNN operations, rebuild TensorFlow with the appropriate compiler flags.\n"
]
}
],
"source": [ "source": [
"from tensorflow.python.client import device_lib\n", "from tensorflow.python.client import device_lib\n",
"print(device_lib.list_local_devices())" "print(device_lib.list_local_devices())"
@ -91,9 +113,17 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": 6,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"CPU\n"
]
}
],
"source": [ "source": [
"MODE = \"GPU\" if \"GPU\" in [k.device_type for k in device_lib.list_local_devices()] else \"CPU\"\n", "MODE = \"GPU\" if \"GPU\" in [k.device_type for k in device_lib.list_local_devices()] else \"CPU\"\n",
"print(MODE)" "print(MODE)"
@ -132,25 +162,24 @@
"\n", "\n",
"```\n", "```\n",
"data_dir\n", "data_dir\n",
"└───sample/\n", "└───train/\n",
"│ └───train/\n", "│ └───absence/\n",
"│ │ └───absence/\n", "│ │ │ absence_t_0001.jpg\n",
"│ │ │ │ absence_t_0001.jpg\n", "│ │ │ absence_t_0002.jpg\n",
"│ │ │ │ absence_t_0002.jpg\n", "│ │ │ ...\n",
"│ │ │ │ ...\n", "│ └───presence/\n",
"│ │ └───presence/\n", "│ │ │ presence_t_0001.jpg\n",
"│ │ │ │ presence_t_0001.jpg\n", "│ │ │ presence_t_0002.jpg\n",
"│ │ │ │ presence_t_0002.jpg\n", "│ │ │ ...\n",
"│ │ │ │ ...\n", "└───validation/\n",
"│ └───validation/\n", "│ └───absence/\n",
"│ │ └───absence/\n", "│ │ │ absence_v_0001.jpg\n",
"│ │ │ │ absence_v_0001.jpg\n", "│ │ │ absence_v_0002.jpg\n",
"│ │ │ │ absence_v_0002.jpg\n", "│ │ │ ...\n",
"│ │ │ │ ...\n", "│ └───presence/\n",
"│ │ └───presence/\n", "│ │ │ presence_v_0001.jpg\n",
"│ │ │ │ presence_v_0001.jpg\n", "│ │ │ presence_v_0002.jpg\n",
"│ │ │ │ presence_v_0002.jpg\n", "│ │ │ ...\n",
"│ │ │ │ ...\n",
"```" "```"
] ]
}, },
@ -158,23 +187,19 @@
"cell_type": "markdown", "cell_type": "markdown",
"metadata": {}, "metadata": {},
"source": [ "source": [
"### Parameter" "### Reading the data"
] ]
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": 7,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"data_dir = 'data/' # data path\n", "data_dir = 'data/' # data path\n",
"\n", "\n",
"# Sample directory path\n", "# Sample directory path\n",
"# That way we can use this algorithm with an other sample if we have mire pictures in the future.\n", "# That way we can use this algorithm with an other sample if we have mire pictures in the future.\n"
"\n",
"N_train = 160\n",
"N_val = 40\n",
"data_dir_sub = data_dir+'sample_%d_Ntrain_%d_Nval' %(N_train, N_val)"
] ]
}, },
{ {
@ -188,23 +213,11 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 1, "execution_count": null,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"ename": "NameError",
"evalue": "name 'kpi' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\u001b[0;32m/var/folders/83/bqq_97cs083fpgkn46bd5k5r0000gn/T/ipykernel_61636/3158407634.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mimg\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mkpi\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload_img\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata_dir_sub\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;34m'/train/presence/presence_t_0001.jpg'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0mimg\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0;31mNameError\u001b[0m: name 'kpi' is not defined"
]
}
],
"source": [ "source": [
"img = kpi.load_img(data_dir_sub+'/train/presence/presence_t_0001.jpg') \n", "img = kpi.load_img(data_dir+'/train/presence/presence_t_0001.jpg') \n",
"img" "img"
] ]
}, },
@ -281,15 +294,15 @@
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"img = kpi.load_img(data_dir_sub+\"/train/presence/presence_t_0001.jpg\")\n", "img = kpi.load_img(data_dir+\"/train/presence/presence_t_0001.jpg\")\n",
"x = kpi.img_to_array(img) \n", "x = kpi.img_to_array(img) \n",
"x_ = np.expand_dims(x, axis=0)\n", "x_ = np.expand_dims(x, axis=0)\n",
"\n", "\n",
"if not(os.path.isdir(data_dir_sub+\"/preprocessing_example\")):\n", "if not(os.path.isdir(data_dir+\"/preprocessing_example\")):\n",
" os.mkdir(data_dir_sub+\"/preprocessing_example\")\n", " os.mkdir(data_dir+\"/preprocessing_example\")\n",
"\n", "\n",
" i = 0\n", " i = 0\n",
" for batch in datagen.flow(x_, batch_size=1,save_to_dir=data_dir_sub+\"/preprocessing_example\", save_prefix='bird', save_format='jpg'):\n", " for batch in datagen.flow(x_, batch_size=1,save_to_dir=data_dir+\"/preprocessing_example\", save_prefix='bird', save_format='jpg'):\n",
" i += 1\n", " i += 1\n",
" if i > 7:\n", " if i > 7:\n",
" break " " break "
@ -330,8 +343,8 @@
], ],
"source": [ "source": [
"X_list=[]\n", "X_list=[]\n",
"for f in os.listdir(data_dir_sub+\"/preprocessing_example\"):\n", "for f in os.listdir(data_dir+\"/preprocessing_example\"):\n",
" X_list.append(kpi.img_to_array(kpi.load_img(data_dir_sub+\"/preprocessing_example/\"+f)))\n", " X_list.append(kpi.img_to_array(kpi.load_img(data_dir+\"/preprocessing_example/\"+f)))\n",
"\n", "\n",
"\n", "\n",
"fig=plt.figure(figsize=(16,8))\n", "fig=plt.figure(figsize=(16,8))\n",
@ -411,8 +424,8 @@
} }
], ],
"source": [ "source": [
"x_0 = kpi.img_to_array(kpi.load_img(data_dir_sub+\"/train/presence/presence_t_0001.jpg\"))\n", "x_0 = kpi.img_to_array(kpi.load_img(data_dir+\"/train/presence/presence_t_0001.jpg\"))\n",
"x_1 = kpi.img_to_array(kpi.load_img(data_dir_sub+\"/train/presence/presence_t_0002.jpg\"))\n", "x_1 = kpi.img_to_array(kpi.load_img(data_dir+\"/train/presence/presence_t_0002.jpg\"))\n",
"x_0.shape, x_1.shape" "x_0.shape, x_1.shape"
] ]
}, },
@ -461,14 +474,14 @@
"# subfolers of 'data/train', and indefinitely generate\n", "# subfolers of 'data/train', and indefinitely generate\n",
"# batches of augmented image data\n", "# batches of augmented image data\n",
"train_generator = train_datagen.flow_from_directory(\n", "train_generator = train_datagen.flow_from_directory(\n",
" data_dir_sub+\"/train/\", # this is the target directory\n", " data_dir+\"/train/\", # this is the target directory\n",
" target_size=(img_width, img_height), \n", " target_size=(img_width, img_height), \n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode='binary') # since we use binary_crossentropy loss, we need binary labels\n", " class_mode='binary') # since we use binary_crossentropy loss, we need binary labels\n",
"\n", "\n",
"# this is a similar generator, for validation data\n", "# this is a similar generator, for validation data\n",
"validation_generator = valid_datagen.flow_from_directory(\n", "validation_generator = valid_datagen.flow_from_directory(\n",
" data_dir_sub+\"/validation/\",\n", " data_dir+\"/validation/\",\n",
" target_size=(img_width, img_height),\n", " target_size=(img_width, img_height),\n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode='binary')" " class_mode='binary')"
@ -840,7 +853,7 @@
"datagen = kpi.ImageDataGenerator(rescale=1. / 255)\n", "datagen = kpi.ImageDataGenerator(rescale=1. / 255)\n",
"\n", "\n",
"generator = datagen.flow_from_directory(\n", "generator = datagen.flow_from_directory(\n",
" data_dir_sub+\"/train\",\n", " data_dir+\"/train\",\n",
" target_size=(img_width, img_height),\n", " target_size=(img_width, img_height),\n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode=None, # this means our generator will only yield batches of data, no labels\n", " class_mode=None, # this means our generator will only yield batches of data, no labels\n",
@ -849,7 +862,7 @@
"\n", "\n",
"\n", "\n",
"generator = datagen.flow_from_directory(\n", "generator = datagen.flow_from_directory(\n",
" data_dir_sub+\"/validation\",\n", " data_dir+\"/validation\",\n",
" target_size=(img_width, img_height),\n", " target_size=(img_width, img_height),\n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode=None,\n", " class_mode=None,\n",
@ -1044,7 +1057,7 @@
} }
], ],
"source": [ "source": [
"model_VGG_fcm.save_weights(data_dir_sub+'/weights_model_VGG_fully_connected_model_%d_epochs_%d_batch_size.h5' %(epochs, batch_size))" "model_VGG_fcm.save_weights(data_dir+'/weights_model_VGG_fully_connected_model_%d_epochs_%d_batch_size.h5' %(epochs, batch_size))"
] ]
}, },
{ {
@ -1113,7 +1126,7 @@
"top_model.add(kl.Dropout(0.5))\n", "top_model.add(kl.Dropout(0.5))\n",
"top_model.add(kl.Dense(1, activation='sigmoid'))\n", "top_model.add(kl.Dense(1, activation='sigmoid'))\n",
"\n", "\n",
"top_model.load_weights(data_dir_sub+'/weights_model_VGG_fully_connected_model_%d_epochs_%d_batch_size.h5' %(epochs, batch_size))\n" "top_model.load_weights(data_dir+'/weights_model_VGG_fully_connected_model_%d_epochs_%d_batch_size.h5' %(epochs, batch_size))\n"
] ]
}, },
{ {
@ -1247,13 +1260,13 @@
"test_datagen = kpi.ImageDataGenerator(rescale=1. / 255)\n", "test_datagen = kpi.ImageDataGenerator(rescale=1. / 255)\n",
"\n", "\n",
"train_generator = train_datagen.flow_from_directory(\n", "train_generator = train_datagen.flow_from_directory(\n",
" data_dir_sub+\"/train/\",\n", " data_dir+\"/train/\",\n",
" target_size=(img_height, img_width),\n", " target_size=(img_height, img_width),\n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode='binary')\n", " class_mode='binary')\n",
"\n", "\n",
"validation_generator = test_datagen.flow_from_directory(\n", "validation_generator = test_datagen.flow_from_directory(\n",
" data_dir_sub+\"/validation/\",\n", " data_dir+\"/validation/\",\n",
" target_size=(img_height, img_width),\n", " target_size=(img_height, img_width),\n",
" batch_size=batch_size,\n", " batch_size=batch_size,\n",
" class_mode='binary')\n" " class_mode='binary')\n"
@ -1390,7 +1403,7 @@
"name": "python", "name": "python",
"nbconvert_exporter": "python", "nbconvert_exporter": "python",
"pygments_lexer": "ipython3", "pygments_lexer": "ipython3",
"version": "3.9.7" "version": "3.7.11"
}, },
"toc": { "toc": {
"nav_menu": {}, "nav_menu": {},

BIN
data/.DS_Store vendored
View file

Binary file not shown.

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/.DS_Store → data/preprocessing_example/.DS_Store vendored
View file

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_17.jpg → data/preprocessing_example/bird_0_17.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 603 KiB

After

Width:  |  Height:  |  Size: 603 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_2061.jpg → data/preprocessing_example/bird_0_2061.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 654 KiB

After

Width:  |  Height:  |  Size: 654 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_5194.jpg → data/preprocessing_example/bird_0_5194.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 586 KiB

After

Width:  |  Height:  |  Size: 586 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_6313.jpg → data/preprocessing_example/bird_0_6313.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 620 KiB

After

Width:  |  Height:  |  Size: 620 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_6504.jpg → data/preprocessing_example/bird_0_6504.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 605 KiB

After

Width:  |  Height:  |  Size: 605 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_6971.jpg → data/preprocessing_example/bird_0_6971.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 626 KiB

After

Width:  |  Height:  |  Size: 626 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_7861.jpg → data/preprocessing_example/bird_0_7861.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 619 KiB

After

Width:  |  Height:  |  Size: 619 KiB

0
data/sample_160_Ntrain_40_Nval/preprocessing_example/bird_0_8433.jpg → data/preprocessing_example/bird_0_8433.jpg
View file

Side by side Swipe Overlay

Before

Width:  |  Height:  |  Size: 672 KiB

After

Width:  |  Height:  |  Size: 672 KiB

BIN
data/sample_160_Ntrain_40_Nval/.DS_Store → data/train/.DS_Store vendored
View file

Binary file not shown.

BIN
data/validation/.DS_Store vendored Normal file
View file

Binary file not shown.