{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# Introduction\n\nDiabetic Retinopathy (DR) is a complication of diabetes, caused by high blood sugar levels damaging the back of the eye (retina). It can cause blindness if left undiagnosed and untreated.\n\nDR is split:\n- Grade 0 and 1: considered as No \"referable\" DR, because 1 is difficult to diagnose (early DR)\n- Grade 2: **background retinopathy** – tiny bulges develop in the blood vessels, which may bleed slightly but do not usually affect your vision\n- Grade 3:**pre-proliferative retinopathy** – more severe and widespread changes affect the blood vessels, including more significant bleeding into the eye\n- Grade 4: **proliferative retinopathy** – scar tissue and new blood vessels, which are weak and bleed easily, develop on the retina; this can result in some loss of vision\n\n*Source: [NHS UK](https://www.nhs.uk/conditions/diabetic-retinopathy/)*\n\nGlobally, the number of people with DR will grow from 126.6 million in 2010 to 191.0 million by 2030.  \n*Source: [10.4103/0301-4738.100542](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3491270/)*\n\n![dr_grades.png](https://www.ophthalytics.com/wp-content/uploads/2021/08/Copy-of-WEBSITE-CONTENT-1536x878.png)\n\n**Source:** Ophthalytics.  \n**Link:** https://www.ophthalytics.com/our-technology/diabetic-retinopathy/\n\nNote we will refer to Diabetic Retinoptahy as DR in the following.\n\nWith:\n- 0 - No DR\n- 1 - Mild\n- 2 - Moderate\n- 3 - Severe\n- 4 - Proliferative DR\n\n## Credits\n\nThis [implementation](https://www.kaggle.com/code/raufmomin/vision-transformer-vit-fine-tuning) was taken from [RAUF MOMIN](https://www.kaggle.com/raufmomin). \n\n## Disclaimer\nThis notebook implements Vision Transformer (ViT) model by Alexey Dosovitskiy et al for image classification, and demonstrates it on the APTOS 2019 Diabetic Retionapathy Classification dataset.\n\nFor from scratch implementation of ViT check out this notebook: <br>\nhttps://www.kaggle.com/raufmomin/vision-transformer-vit-from-scratch\n\nResearch Paper: https://arxiv.org/pdf/2010.11929.pdf <br>\nGithub (Official) Link: https://github.com/google-research/vision_transformer <br>\nGithub (Keras) Link: https://github.com/faustomorales/vit-keras","metadata":{}},{"cell_type":"markdown","source":"#### Highlights of this notebook:\n1. Pre-trained Vision Transformer (vit_b32) on imagenet21k dataset\n2. Label Smoothing of 0.3\n3. Custom data augmentation for ImageDataGenerator\n4. RectifiedAdam Optimizer\n5. Grad CAM\n\n# ViT Architecture\n\n![figure1.png](attachment:figure1.png)","metadata":{},"attachments":{"figure1.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"# Libraries and Configurations","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nimport tensorflow as tf\nimport tensorflow_addons as tfa\nimport glob, warnings\nimport matplotlib.pyplot as plt\nfrom sklearn.metrics import confusion_matrix, classification_report\nimport seaborn as sns\n\nwarnings.filterwarnings('ignore')\nprint('TensorFlow Version ' + tf.__version__)","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-10-18T16:40:21.758224Z","iopub.execute_input":"2022-10-18T16:40:21.75865Z","iopub.status.idle":"2022-10-18T16:40:29.301635Z","shell.execute_reply.started":"2022-10-18T16:40:21.758562Z","shell.execute_reply":"2022-10-18T16:40:29.300594Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"IMAGE_SIZE = 224\nBATCH_SIZE = 16\nEPOCHS = 7\n\nTRAIN_PATH = '../input/aptos2019-blindness-detection/train_images'\nTEST_PATH = '../input/aptos2019-blindness-detection/test_images'\n\nDF_TRAIN = pd.read_csv('../input/aptos2019-blindness-detection/train.csv', dtype='str')\nDF_TRAIN['image_path'] = DF_TRAIN[\"id_code\"] + \".png\" \nTEST_IMAGES = glob.glob(TEST_PATH + '/*.png')\nDF_TEST = pd.DataFrame(TEST_IMAGES, columns = ['image_path'])\n\nclasses = {0 : \"No DR\",\n           1 : \"Mild\",\n           2 : \"Moderate\",\n           3 : \"Severe\",\n           4 : \"Proliferative\"}","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:47:12.70358Z","iopub.execute_input":"2022-10-18T16:47:12.704346Z","iopub.status.idle":"2022-10-18T16:47:12.744961Z","shell.execute_reply.started":"2022-10-18T16:47:12.704307Z","shell.execute_reply":"2022-10-18T16:47:12.743824Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"DF_TRAIN.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:47:14.445698Z","iopub.execute_input":"2022-10-18T16:47:14.446091Z","iopub.status.idle":"2022-10-18T16:47:14.458485Z","shell.execute_reply.started":"2022-10-18T16:47:14.446056Z","shell.execute_reply":"2022-10-18T16:47:14.457346Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"DF_TEST.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:47:26.322017Z","iopub.execute_input":"2022-10-18T16:47:26.322376Z","iopub.status.idle":"2022-10-18T16:47:26.332058Z","shell.execute_reply.started":"2022-10-18T16:47:26.322339Z","shell.execute_reply":"2022-10-18T16:47:26.330912Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Augmentations","metadata":{}},{"cell_type":"code","source":"def data_augment(image):\n    p_spatial = tf.random.uniform([], 0, 1.0, dtype = tf.float32)\n    #p_rotate = tf.random.uniform([], 0, 1.0, dtype = tf.float32)\n    p_pixel_1 = tf.random.uniform([], 0, 1.0, dtype = tf.float32)\n    p_pixel_2 = tf.random.uniform([], 0, 1.0, dtype = tf.float32)\n    p_pixel_3 = tf.random.uniform([], 0, 1.0, dtype = tf.float32)\n    \n    # Flips\n    #image = tf.image.random_flip_left_right(image)\n    #image = tf.image.random_flip_up_down(image)\n    \n    if p_spatial > .75:\n        image = tf.image.transpose(image)\n        \n    # Rotates\n    #if p_rotate > .75:\n        #image = tf.image.rot90(image, k = 3) # rotate 270º\n    #elif p_rotate > .5:\n        #image = tf.image.rot90(image, k = 2) # rotate 180º\n    #elif p_rotate > .25:\n        #image = tf.image.rot90(image, k = 1) # rotate 90º\n        \n    # Pixel-level transforms\n    if p_pixel_1 >= .4:\n        image = tf.image.random_saturation(image, lower = .7, upper = 1.3)\n    if p_pixel_2 >= .4:\n        image = tf.image.random_contrast(image, lower = .8, upper = 1.2)\n    if p_pixel_3 >= .4:\n        image = tf.image.random_brightness(image, max_delta = .1)\n        \n    return image","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:56:44.540377Z","iopub.execute_input":"2022-10-18T16:56:44.54078Z","iopub.status.idle":"2022-10-18T16:56:44.549916Z","shell.execute_reply.started":"2022-10-18T16:56:44.540747Z","shell.execute_reply":"2022-10-18T16:56:44.548625Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Generator","metadata":{}},{"cell_type":"code","source":"datagen = tf.keras.preprocessing.image.ImageDataGenerator(rescale = 1./255,\n                                                          samplewise_center = True,\n                                                          samplewise_std_normalization = True,\n                                                          validation_split = 0.2,\n                                                          preprocessing_function = data_augment)\n\ntrain_gen = datagen.flow_from_dataframe(dataframe = DF_TRAIN,\n                                        directory = TRAIN_PATH,\n                                        x_col = 'image_path',\n                                        y_col = 'diagnosis',\n                                        subset = 'training',\n                                        batch_size = BATCH_SIZE,\n                                        seed = 1,\n                                        color_mode = 'rgb',\n                                        shuffle = True,\n                                        class_mode = 'categorical',\n                                        target_size = (IMAGE_SIZE, IMAGE_SIZE))\n\nvalid_gen = datagen.flow_from_dataframe(dataframe = DF_TRAIN,\n                                        directory = TRAIN_PATH,\n                                        x_col = 'image_path',\n                                        y_col = 'diagnosis',\n                                        subset = 'validation',\n                                        batch_size = BATCH_SIZE,\n                                        seed = 1,\n                                        color_mode = 'rgb',\n                                        shuffle = False,\n                                        class_mode = 'categorical',\n                                        target_size = (IMAGE_SIZE, IMAGE_SIZE))\n\ntest_gen = datagen.flow_from_dataframe(dataframe = DF_TEST,\n                                       x_col = 'image_path',\n                                       y_col = None,\n                                       batch_size = BATCH_SIZE,\n                                       seed = 1,\n                                       color_mode = 'rgb',\n                                       shuffle = False,\n                                       class_mode = None,\n                                       target_size = (IMAGE_SIZE, IMAGE_SIZE))","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:57:31.482034Z","iopub.execute_input":"2022-10-18T16:57:31.482445Z","iopub.status.idle":"2022-10-18T16:57:33.821132Z","shell.execute_reply.started":"2022-10-18T16:57:31.482411Z","shell.execute_reply":"2022-10-18T16:57:33.819999Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"images = [train_gen[0][0][i] for i in range(16)]\nfig, axes = plt.subplots(3, 5, figsize = (10, 10))\n\naxes = axes.flatten()\n\nfor img, ax in zip(images, axes):\n    ax.imshow(img.reshape(IMAGE_SIZE, IMAGE_SIZE, 3))\n    ax.axis('off')\n\nplt.tight_layout()\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:57:55.443682Z","iopub.execute_input":"2022-10-18T16:57:55.444283Z","iopub.status.idle":"2022-10-18T16:58:27.952991Z","shell.execute_reply.started":"2022-10-18T16:57:55.444235Z","shell.execute_reply":"2022-10-18T16:58:27.951934Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Building the Model","metadata":{}},{"cell_type":"code","source":"!pip install --quiet vit-keras\n\nfrom vit_keras import vit","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:59:17.632382Z","iopub.execute_input":"2022-10-18T16:59:17.63276Z","iopub.status.idle":"2022-10-18T16:59:30.114096Z","shell.execute_reply.started":"2022-10-18T16:59:17.632728Z","shell.execute_reply":"2022-10-18T16:59:30.112772Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## 1. ViT B32 Model","metadata":{}},{"cell_type":"code","source":"vit_model = vit.vit_b32(\n        image_size = IMAGE_SIZE,\n        activation = 'softmax',\n        pretrained = True,\n        include_top = False,\n        pretrained_top = False,\n        classes = 5)","metadata":{"execution":{"iopub.status.busy":"2022-10-18T16:59:40.117216Z","iopub.execute_input":"2022-10-18T16:59:40.117622Z","iopub.status.idle":"2022-10-18T17:00:44.852646Z","shell.execute_reply.started":"2022-10-18T16:59:40.117586Z","shell.execute_reply":"2022-10-18T17:00:44.851681Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Visualizing Attention Maps of Sample Test Image","metadata":{}},{"cell_type":"code","source":"from vit_keras import visualize\n\nx = test_gen.next()\nimage = x[0]\n\nattention_map = visualize.attention_map(model = vit_model, image = image)\n\n# Plot results\nfig, (ax1, ax2) = plt.subplots(ncols = 2)\nax1.axis('off')\nax2.axis('off')\nax1.set_title('Original')\nax2.set_title('Attention Map')\n_ = ax1.imshow(image)\n_ = ax2.imshow(attention_map)","metadata":{"execution":{"iopub.status.busy":"2022-10-18T17:00:49.110779Z","iopub.execute_input":"2022-10-18T17:00:49.111178Z","iopub.status.idle":"2022-10-18T17:00:54.294442Z","shell.execute_reply.started":"2022-10-18T17:00:49.11114Z","shell.execute_reply":"2022-10-18T17:00:54.293447Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## 2. Fine-tuning the Model","metadata":{}},{"cell_type":"code","source":"model = tf.keras.Sequential([\n        vit_model,\n        tf.keras.layers.Flatten(),\n        tf.keras.layers.BatchNormalization(),\n        tf.keras.layers.Dense(11, activation = tfa.activations.gelu),\n        tf.keras.layers.BatchNormalization(),\n        tf.keras.layers.Dense(5, 'softmax')\n    ],\n    name = 'vision_transformer')\n\nmodel.summary()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T17:01:30.394479Z","iopub.execute_input":"2022-10-18T17:01:30.394906Z","iopub.status.idle":"2022-10-18T17:01:31.97225Z","shell.execute_reply.started":"2022-10-18T17:01:30.394872Z","shell.execute_reply":"2022-10-18T17:01:31.970744Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model_b = tf.keras.Sequential([\n        vit_model,\n        tf.keras.layers.LayerNormalization(),\n        tf.keras.layers.Flatten(),\n        tf.keras.layers.BatchNormalization(),\n        tf.keras.layers.Dense(11, activation = tfa.activations.gelu),\n        tf.keras.layers.BatchNormalization(),\n        tf.keras.layers.Dense(5, 'softmax')\n    ],\n    name = 'vision_transformer')\n\nmodel_b.summary()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T20:08:23.997627Z","iopub.execute_input":"2022-10-18T20:08:23.998253Z","iopub.status.idle":"2022-10-18T20:08:25.98096Z","shell.execute_reply.started":"2022-10-18T20:08:23.998217Z","shell.execute_reply":"2022-10-18T20:08:25.979814Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Training the Model","metadata":{}},{"cell_type":"code","source":"learning_rate = 1e-4\n\noptimizer = tfa.optimizers.RectifiedAdam(learning_rate = learning_rate)\n\nmodel.compile(optimizer = optimizer, \n              loss = tf.keras.losses.CategoricalCrossentropy(label_smoothing = 0.2), \n              metrics = ['accuracy'])\n\nSTEP_SIZE_TRAIN = train_gen.n // train_gen.batch_size\nSTEP_SIZE_VALID = valid_gen.n // valid_gen.batch_size\n\nreduce_lr = tf.keras.callbacks.ReduceLROnPlateau(monitor = 'val_accuracy',\n                                                 factor = 0.2,\n                                                 patience = 2,\n                                                 verbose = 1,\n                                                 min_delta = 1e-4,\n                                                 min_lr = 1e-6,\n                                                 mode = 'max')\n\nearlystopping = tf.keras.callbacks.EarlyStopping(monitor = 'val_accuracy',\n                                                 min_delta = 1e-4,\n                                                 patience = 5,\n                                                 mode = 'max',\n                                                 restore_best_weights = True,\n                                                 verbose = 1)\n\ncheckpointer = tf.keras.callbacks.ModelCheckpoint(filepath = './model.hdf5',\n                                                  monitor = 'val_accuracy', \n                                                  verbose = 1, \n                                                  save_best_only = True,\n                                                  save_weights_only = True,\n                                                  mode = 'max')\n\ncallbacks = [earlystopping, reduce_lr, checkpointer]\n\nmodel.fit(x = train_gen,\n          steps_per_epoch = STEP_SIZE_TRAIN,\n          validation_data = valid_gen,\n          validation_steps = STEP_SIZE_VALID,\n          epochs = EPOCHS,\n          callbacks = callbacks)\n\n#model.save('model.h5', save_weights_only = True) #NO NEED AS THE WEIGHTS ARE ALREADY SAVED","metadata":{"execution":{"iopub.status.busy":"2022-10-18T17:06:44.777077Z","iopub.execute_input":"2022-10-18T17:06:44.777524Z","iopub.status.idle":"2022-10-18T19:12:53.631871Z","shell.execute_reply.started":"2022-10-18T17:06:44.777489Z","shell.execute_reply":"2022-10-18T19:12:53.621921Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.load_weights(\"./model.hdf5\")","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:38:54.577415Z","iopub.execute_input":"2022-10-18T19:38:54.577887Z","iopub.status.idle":"2022-10-18T19:38:55.153414Z","shell.execute_reply.started":"2022-10-18T19:38:54.577847Z","shell.execute_reply":"2022-10-18T19:38:55.152217Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:39:02.965625Z","iopub.execute_input":"2022-10-18T19:39:02.966053Z","iopub.status.idle":"2022-10-18T19:39:02.991028Z","shell.execute_reply.started":"2022-10-18T19:39:02.966008Z","shell.execute_reply":"2022-10-18T19:39:02.988407Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model.get_layer(\"vit-b32\").summary()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:50:07.757863Z","iopub.execute_input":"2022-10-18T19:50:07.7583Z","iopub.status.idle":"2022-10-18T19:50:07.781544Z","shell.execute_reply.started":"2022-10-18T19:50:07.758263Z","shell.execute_reply":"2022-10-18T19:50:07.780046Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Results","metadata":{}},{"cell_type":"code","source":"predicted_classes = np.argmax(model.predict(valid_gen, steps = valid_gen.n // valid_gen.batch_size + 1), axis = 1)\ntrue_classes = valid_gen.classes\nclass_labels = list(valid_gen.class_indices.keys())  \n\nconfusionmatrix = confusion_matrix(true_classes, predicted_classes)\nplt.figure(figsize = (16, 16))\nsns.heatmap(confusionmatrix, cmap = 'Blues', annot = True, cbar = True)\n\nprint(classification_report(true_classes, predicted_classes))","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:14:24.533268Z","iopub.execute_input":"2022-10-18T19:14:24.53402Z","iopub.status.idle":"2022-10-18T19:16:36.702841Z","shell.execute_reply.started":"2022-10-18T19:14:24.533969Z","shell.execute_reply":"2022-10-18T19:16:36.701813Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Creates a confusion matrix\ncm = confusion_matrix(true_classes, predicted_classes) \n\n# Transform to df for easier plotting\ncm_df = pd.DataFrame(cm,\n                     index = [\"No DR\", \"Mild\", \"Moderate\", \"Severe\", \"Proliferative DR\"], \n                     columns = [\"No DR\", \"Mild\", \"Moderate\", \"Severe\", \"Proliferative DR\"])\n\nplt.figure(figsize=(6,4))\nsns.heatmap(cm_df, square=True, annot=True, cmap=\"Greens\", fmt='d', cbar=False)\nplt.title('Fine tuned ViT Model Confusion Matrix')\nplt.ylabel('True label')\nplt.xlabel('Predicted label')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:23:48.776353Z","iopub.execute_input":"2022-10-18T19:23:48.776889Z","iopub.status.idle":"2022-10-18T19:23:49.071301Z","shell.execute_reply.started":"2022-10-18T19:23:48.776826Z","shell.execute_reply":"2022-10-18T19:23:49.070404Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Investigate Model's Predictions","metadata":{}},{"cell_type":"markdown","source":"## Grad CAM Algorithm\n\nSource: https://www.kaggle.com/code/easara/covid19-detection-with-vit-and-heatmap/edit","metadata":{}},{"cell_type":"code","source":"def get_img_array(img):\n    \n    # `array` is a float32 Numpy array of shape (299, 299, 3)\n    array = keras.preprocessing.image.img_to_array(img)\n    # We add a dimension to transform our array into a \"batch\"\n    # of size (1, 299, 299, 3)\n    array = np.expand_dims(array, axis=0)\n    return array","metadata":{"execution":{"iopub.status.busy":"2022-10-18T20:02:20.732932Z","iopub.execute_input":"2022-10-18T20:02:20.733396Z","iopub.status.idle":"2022-10-18T20:02:20.739179Z","shell.execute_reply.started":"2022-10-18T20:02:20.733359Z","shell.execute_reply":"2022-10-18T20:02:20.737958Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def gradcam_heatmap(img_array, model, last_conv_layer_name, pred_index=None):\n    # First, we create a model that maps the input image to the activations\n    # of the last conv layer as well as the output predictions\n    grad_model = tf.keras.models.Model(\n        [model.input], [model.get_layer(last_conv_layer_name).output,  model.output]\n    )\n    \n    # Then, we compute the gradient of the top predicted class for our input image\n    # with respect to the activations of the last conv layer\n    with tf.GradientTape() as tape:\n        last_conv_layer_output, preds = grad_model(img_array)\n        if pred_index is None:\n            pred_index = tf.argmax(preds[0])\n        class_channel = preds[:, pred_index]\n        \n        \n    # This is the gradient of the output neuron (top predicted or chosen)\n    # with regard to the output feature map of the last conv layer\n    grads = tape.gradient(class_channel, last_conv_layer_output)\n\n    # This is a vector where each entry is the mean intensity of the gradient\n    # over a specific feature map channel\n    pooled_grads = tf.reduce_mean(grads, axis=(0, 1))\n    # We multiply each channel in the feature map array\n    # by \"how important this channel is\" with regard to the top predicted class\n    # then sum all the channels to obtain the heatmap class activation\n    last_conv_layer_output = last_conv_layer_output#[0]\n    #print(np.expand_dims(last_conv_layer_output,axis=0))\n    #print(pooled_grads[..., tf.newaxis])\n    heatmap = last_conv_layer_output @ pooled_grads[..., tf.newaxis]\n    heatmap = tf.squeeze(heatmap)\n    \n    # For visualization purpose, we will also normalize the heatmap between 0 & 1\n    heatmap = tf.maximum(heatmap, 0) / tf.math.reduce_max(heatmap)\n    return heatmap.numpy()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T20:02:21.05969Z","iopub.execute_input":"2022-10-18T20:02:21.060075Z","iopub.status.idle":"2022-10-18T20:02:21.068841Z","shell.execute_reply.started":"2022-10-18T20:02:21.06004Z","shell.execute_reply":"2022-10-18T20:02:21.067401Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def display_gradcam(img, heatmap, cam_path=\"cam.jpg\", alpha=0.4,preds=[0,0,0,0], plot=None):\n\n    # Rescale heatmap to a range 0-255\n    heatmap = np.uint8(255 * heatmap)\n\n    # Use jet colormap to colorize heatmap\n    jet = cm.get_cmap(\"jet\")\n\n    # Use RGB values of the colormap\n    jet_colors = jet(np.arange(256))[:, :3]\n    jet_heatmap = jet_colors[heatmap]\n\n    # Create an image with RGB colorized heatmap\n    jet_heatmap = keras.preprocessing.image.array_to_img(jet_heatmap)\n    jet_heatmap = jet_heatmap.resize((img.shape[1], img.shape[0]))\n    jet_heatmap = keras.preprocessing.image.img_to_array(jet_heatmap)\n\n    # Superimpose the heatmap on original image\n    superimposed_img = jet_heatmap * alpha + img\n    superimposed_img = keras.preprocessing.image.array_to_img(superimposed_img)\n\n    # Save the superimposed image\n    #superimposed_img.save(cam_path)\n\n    # Display Grad CAM\n    #display(Image(cam_path))\n    #plt.figure(figsize=(8,8))\n    plot.imshow(superimposed_img)\n    plot.set(title =\n        \"No DR: \\\n        {:.3f}\\nMild: \\\n        {:.3f}\\nModerate: \\\n        {:.3f}\\nSevere: \\\n        {:.3f}\\nProliferative: \\\n        {:.3f}\".format(preds[0], \\\n                    preds[1], \\\n                    preds[2], \\\n                    preds[3],\n                    preds[4])\n    )\n    plot.axis('off')","metadata":{"execution":{"iopub.status.busy":"2022-10-18T20:02:25.312453Z","iopub.execute_input":"2022-10-18T20:02:25.313173Z","iopub.status.idle":"2022-10-18T20:02:25.324279Z","shell.execute_reply.started":"2022-10-18T20:02:25.313131Z","shell.execute_reply":"2022-10-18T20:02:25.322647Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import tensorflow as tf\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\n\ntest_image = next(iter(valid_gen))[0][5]\n# Prepare image\nimg_array = get_img_array(test_image)\n\nlast_conv_layer_name = 'layer_normalization'\n# Remove last layer's softmax\nmodel.layers[-1].activation = None\n# Print what the top predicted class is\npreds = model.predict(img_array)\nprint(\"Predicted:\\n\" +\"No DR: \\\n    {p1}\\nMild: {p2}\\nModerate: \\\n    {p3}\\nSevere: \\\n    {p4} \\nProliferative: {p5}\".format(p1=preds[0][0], \\\n                                            p2=preds[0][1],p3=preds[0][2],p4=preds[0][3],p5=preds[0][4]))\n# Generate class activation heatmap\nheatmap = gradcam_heatmap(img_array, model, last_conv_layer_name)\n\nheatmap = np.reshape(heatmap, (12,12))\n# Display heatmap\nplt.matshow(heatmap)\n\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T20:15:15.962468Z","iopub.execute_input":"2022-10-18T20:15:15.962894Z","iopub.status.idle":"2022-10-18T20:15:18.236748Z","shell.execute_reply.started":"2022-10-18T20:15:15.962859Z","shell.execute_reply":"2022-10-18T20:15:18.23485Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"filenames_path = valid_gen.filenames\n\nindex_diff, index_same = [], []\nfor i, t, p in zip(range(len(true_classes)), true_classes, predicted_classes):\n    if  t != p:\n        index_diff.append(i)\n    else:\n        index_same.append(i)\n        \npredictions = [int(p.item(0)) for p in predicted_classes]\nfilenamepath_diff = [filenames_path[i] for i in index_diff]\nground_truth_diff = [true_classes[i] for i in index_diff]\npredictions_diff = [predictions[i] for i in index_diff]\ndiff_pred_df = {\"Ground_Truth\": ground_truth_diff,\n                 \"Prediction\": predictions_diff,\n                 \"Filename_Path\": filenamepath_diff}\ndf_diff = pd.DataFrame(diff_pred_df)\ndf_diff.to_csv(\"vitB32_finetuned_failures.csv\")\n\nfilenamepath_same = [filenames_path[i] for i in index_same]\nground_truth_same = [true_classes[i] for i in index_same]\npredictions_same = [predictions[i] for i in index_same]\nsame_pred_df = {\"Ground_Truth\": ground_truth_same,\n                 \"Prediction\": predictions_same,\n                 \"Filename_Path\": filenamepath_same}\ndf_same = pd.DataFrame(same_pred_df)\ndf_same.to_csv(\"vitB32_finetuned_successes.csv\") ","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:28:11.349996Z","iopub.execute_input":"2022-10-18T19:28:11.350411Z","iopub.status.idle":"2022-10-18T19:28:11.397575Z","shell.execute_reply.started":"2022-10-18T19:28:11.350375Z","shell.execute_reply":"2022-10-18T19:28:11.396592Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# The model made ~22% errors over the val set\ndf_diff.shape, df_same.shape","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:28:29.466877Z","iopub.execute_input":"2022-10-18T19:28:29.467255Z","iopub.status.idle":"2022-10-18T19:28:29.473654Z","shell.execute_reply.started":"2022-10-18T19:28:29.467224Z","shell.execute_reply":"2022-10-18T19:28:29.472815Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_diff.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:29:32.959654Z","iopub.execute_input":"2022-10-18T19:29:32.960285Z","iopub.status.idle":"2022-10-18T19:29:32.984117Z","shell.execute_reply.started":"2022-10-18T19:29:32.960227Z","shell.execute_reply":"2022-10-18T19:29:32.982813Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_same.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:29:40.384479Z","iopub.execute_input":"2022-10-18T19:29:40.384895Z","iopub.status.idle":"2022-10-18T19:29:40.395512Z","shell.execute_reply.started":"2022-10-18T19:29:40.384862Z","shell.execute_reply":"2022-10-18T19:29:40.394488Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Example of the model's failure\n\nThe class with the most failures is Moderate being considered as Mild (34 misclassifications) ","metadata":{}},{"cell_type":"code","source":"Moderate = df_diff[df_diff[\"Ground_Truth\"] == 2]\n# 2:Moderate confused with 0: no DR and 1: Mild\nModerate[(Moderate[\"Prediction\"] == 0) | (Moderate[\"Prediction\"] == 1)].head(10)","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:33:35.91987Z","iopub.execute_input":"2022-10-18T19:33:35.920293Z","iopub.status.idle":"2022-10-18T19:33:35.939384Z","shell.execute_reply.started":"2022-10-18T19:33:35.920262Z","shell.execute_reply":"2022-10-18T19:33:35.938192Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"Moderate.reset_index(drop=True, inplace=True)\nModerate.head(-2)","metadata":{"execution":{"iopub.status.busy":"2022-10-18T19:33:49.126132Z","iopub.execute_input":"2022-10-18T19:33:49.126524Z","iopub.status.idle":"2022-10-18T19:33:49.145358Z","shell.execute_reply.started":"2022-10-18T19:33:49.126489Z","shell.execute_reply":"2022-10-18T19:33:49.144289Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Resources\n\nFor more details, check :\n- [Paper](https://arxiv.org/pdf/2010.11929.pdf)\n- [Implement ViT from scratch version](https://www.kaggle.com/code/basu369victor/covid19-detection-with-vit-and-heatmap) : easier to implement Grad CAM with, as you can directly access Layer Normalization\n- Check out the [Hybrid-Effcient-SwinTransfomer version](https://github.com/elateifsara/devoxxMA22/blob/main/devoxxma22-inspect-your-model.ipynb)","metadata":{}},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}