View on TensorFlow.org

Run in Google Colab

View source on GitHub

Download notebook

Given an image like the example below, our goal is to generate a caption such as "a surfer riding on a wave".

Image Source; License: Public Domain

To accomplish this, you'll use an attention-based model, which enables us to see what parts of the image the model focuses on as it generates a caption.

The model architecture is similar to Show, Attend and Tell: Neural Image Caption Generation with Visual Attention.

This notebook is an end-to-end example. When you run the notebook, it downloads the MS-COCO dataset, preprocesses and caches a subset of images using Inception V3, trains an encoder-decoder model, and generates captions on new images using the trained model.

In this example, you will train a model on a relatively small amount of data—the first 30,000 captions for about 20,000 images (because there are multiple captions per image in the dataset).

from __future__ import absolute_import, division, print_function, unicode_literals

import tensorflow as tf

# You'll generate plots of attention in order to see which parts of an image
# our model focuses on during captioning
import matplotlib.pyplot as plt

# Scikit-learn includes many helpful utilities
from sklearn.model_selection import train_test_split
from sklearn.utils import shuffle

import re
import numpy as np
import os
import time
import json
from glob import glob
from PIL import Image
import pickle

Download and prepare the MS-COCO dataset

You will use the MS-COCO dataset to train our model. The dataset contains over 82,000 images, each of which has at least 5 different caption annotations. The code below downloads and extracts the dataset automatically.

Caution: large download ahead. You'll use the training set, which is a 13GB file.

annotation_zip = tf.keras.utils.get_file('captions.zip',
                                          cache_subdir=os.path.abspath('.'),
                                          origin = 'http://images.cocodataset.org/annotations/annotations_trainval2014.zip',
                                          extract = True)
annotation_file = os.path.dirname(annotation_zip)+'/annotations/captions_train2014.json'

name_of_zip = 'train2014.zip'
if not os.path.exists(os.path.abspath('.') + '/' + name_of_zip):
  image_zip = tf.keras.utils.get_file(name_of_zip,
                                      cache_subdir=os.path.abspath('.'),
                                      origin = 'http://images.cocodataset.org/zips/train2014.zip',
                                      extract = True)
  PATH = os.path.dirname(image_zip)+'/train2014/'
else:
  PATH = os.path.abspath('.')+'/train2014/'

Downloading data from http://images.cocodataset.org/annotations/annotations_trainval2014.zip
252878848/252872794 [==============================] - 7s 0us/step
Downloading data from http://images.cocodataset.org/zips/train2014.zip
13510574080/13510573713 [==============================] - 360s 0us/step

Optional: limit the size of the training set

To speed up training for this tutorial, you'll use a subset of 30,000 captions and their corresponding images to train our model. Choosing to use more data would result in improved captioning quality.

# Read the json file
with open(annotation_file, 'r') as f:
    annotations = json.load(f)

# Store captions and image names in vectors
all_captions = []
all_img_name_vector = []

for annot in annotations['annotations']:
    caption = '<start> ' + annot['caption'] + ' <end>'
    image_id = annot['image_id']
    full_coco_image_path = PATH + 'COCO_train2014_' + '%012d.jpg' % (image_id)

    all_img_name_vector.append(full_coco_image_path)
    all_captions.append(caption)

# Shuffle captions and image_names together
# Set a random state
train_captions, img_name_vector = shuffle(all_captions,
                                          all_img_name_vector,
                                          random_state=1)

# Select the first 30000 captions from the shuffled set
num_examples = 30000
train_captions = train_captions[:num_examples]
img_name_vector = img_name_vector[:num_examples]

len(train_captions), len(all_captions)

(30000, 414113)

Preprocess the images using InceptionV3

Next, you will use InceptionV3 (which is pretrained on Imagenet) to classify each image. You will extract features from the last convolutional layer.

First, you will convert the images into InceptionV3's expected format by:

• Resizing the image to 299px by 299px
• Preprocess the images using the preprocess_input method to normalize the image so that it contains pixels in the range of -1 to 1, which matches the format of the images used to train InceptionV3.

def load_image(image_path):
    img = tf.io.read_file(image_path)
    img = tf.image.decode_jpeg(img, channels=3)
    img = tf.image.resize(img, (299, 299))
    img = tf.keras.applications.inception_v3.preprocess_input(img)
    return img, image_path

Initialize InceptionV3 and load the pretrained Imagenet weights

Now you'll create a tf.keras model where the output layer is the last convolutional layer in the InceptionV3 architecture. The shape of the output of this layer is 8x8x2048. You use the last convolutional layer because you are using attention in this example. You don't perform this initialization during training because it could become a bottleneck.

• You forward each image through the network and store the resulting vector in a dictionary (image_name --> feature_vector).
• After all the images are passed through the network, you pickle the dictionary and save it to disk.

image_model = tf.keras.applications.InceptionV3(include_top=False,
                                                weights='imagenet')
new_input = image_model.input
hidden_layer = image_model.layers[-1].output

image_features_extract_model = tf.keras.Model(new_input, hidden_layer)

Downloading data from https://github.com/fchollet/deep-learning-models/releases/download/v0.5/inception_v3_weights_tf_dim_ordering_tf_kernels_notop.h5
87916544/87910968 [==============================] - 3s 0us/step

Caching the features extracted from InceptionV3

You will pre-process each image with InceptionV3 and cache the output to disk. Caching the output in RAM would be faster but also memory intensive, requiring 8 * 8 * 2048 floats per image. At the time of writing, this exceeds the memory limitations of Colab (currently 12GB of memory).

Performance could be improved with a more sophisticated caching strategy (for example, by sharding the images to reduce random access disk I/O), but that would require more code.

The caching will take about 10 minutes to run in Colab with a GPU. If you'd like to see a progress bar, you can:

1. install tqdm:

   !pip install -q tqdm

2. Import tqdm:

   from tqdm import tqdm

3. Change the following line:

   for img, path in image_dataset:

   to:

   for img, path in tqdm(image_dataset):

# Get unique images
encode_train = sorted(set(img_name_vector))

# Feel free to change batch_size according to your system configuration
image_dataset = tf.data.Dataset.from_tensor_slices(encode_train)
image_dataset = image_dataset.map(
  load_image, num_parallel_calls=tf.data.experimental.AUTOTUNE).batch(16)

for img, path in image_dataset:
  batch_features = image_features_extract_model(img)
  batch_features = tf.reshape(batch_features,
                              (batch_features.shape[0], -1, batch_features.shape[3]))

  for bf, p in zip(batch_features, path):
    path_of_feature = p.numpy().decode("utf-8")
    np.save(path_of_feature, bf.numpy())

Preprocess and tokenize the captions

• First, you'll tokenize the captions (for example, by splitting on spaces). This gives us a vocabulary of all of the unique words in the data (for example, "surfing", "football", and so on).
• Next, you'll limit the vocabulary size to the top 5,000 words (to save memory). You'll replace all other words with the token "UNK" (unknown).
• You then create word-to-index and index-to-word mappings.
• Finally, you pad all sequences to be the same length as the longest one.

# Find the maximum length of any caption in our dataset
def calc_max_length(tensor):
    return max(len(t) for t in tensor)

# Choose the top 5000 words from the vocabulary
top_k = 5000
tokenizer = tf.keras.preprocessing.text.Tokenizer(num_words=top_k,
                                                  oov_token="<unk>",
                                                  filters='!"#$%&()*+.,-/:;=?@[\]^_`{|}~ ')
tokenizer.fit_on_texts(train_captions)
train_seqs = tokenizer.texts_to_sequences(train_captions)

tokenizer.word_index['<pad>'] = 0
tokenizer.index_word[0] = '<pad>'

# Create the tokenized vectors
train_seqs = tokenizer.texts_to_sequences(train_captions)

# Pad each vector to the max_length of the captions
# If you do not provide a max_length value, pad_sequences calculates it automatically
cap_vector = tf.keras.preprocessing.sequence.pad_sequences(train_seqs, padding='post')

# Calculates the max_length, which is used to store the attention weights
max_length = calc_max_length(train_seqs)

Split the data into training and testing

# Create training and validation sets using an 80-20 split
img_name_train, img_name_val, cap_train, cap_val = train_test_split(img_name_vector,
                                                                    cap_vector,
                                                                    test_size=0.2,
                                                                    random_state=0)

len(img_name_train), len(cap_train), len(img_name_val), len(cap_val)

(24000, 24000, 6000, 6000)

Create a tf.data dataset for training

Our images and captions are ready! Next, let's create a tf.data dataset to use for training our model.

# Feel free to change these parameters according to your system's configuration

BATCH_SIZE = 64
BUFFER_SIZE = 1000
embedding_dim = 256
units = 512
vocab_size = len(tokenizer.word_index) + 1
num_steps = len(img_name_train) // BATCH_SIZE
# Shape of the vector extracted from InceptionV3 is (64, 2048)
# These two variables represent that vector shape
features_shape = 2048
attention_features_shape = 64

# Load the numpy files
def map_func(img_name, cap):
  img_tensor = np.load(img_name.decode('utf-8')+'.npy')
  return img_tensor, cap

dataset = tf.data.Dataset.from_tensor_slices((img_name_train, cap_train))

# Use map to load the numpy files in parallel
dataset = dataset.map(lambda item1, item2: tf.numpy_function(
          map_func, [item1, item2], [tf.float32, tf.int32]),
          num_parallel_calls=tf.data.experimental.AUTOTUNE)

# Shuffle and batch
dataset = dataset.shuffle(BUFFER_SIZE).batch(BATCH_SIZE)
dataset = dataset.prefetch(buffer_size=tf.data.experimental.AUTOTUNE)

Model

Fun fact: the decoder below is identical to the one in the example for Neural Machine Translation with Attention.

The model architecture is inspired by the Show, Attend and Tell paper.

• In this example, you extract the features from the lower convolutional layer of InceptionV3 giving us a vector of shape (8, 8, 2048).
• You squash that to a shape of (64, 2048).
• This vector is then passed through the CNN Encoder (which consists of a single Fully connected layer).
• The RNN (here GRU) attends over the image to predict the next word.

class BahdanauAttention(tf.keras.Model):
  def __init__(self, units):
    super(BahdanauAttention, self).__init__()
    self.W1 = tf.keras.layers.Dense(units)
    self.W2 = tf.keras.layers.Dense(units)
    self.V = tf.keras.layers.Dense(1)

  def call(self, features, hidden):
    # features(CNN_encoder output) shape == (batch_size, 64, embedding_dim)

    # hidden shape == (batch_size, hidden_size)
    # hidden_with_time_axis shape == (batch_size, 1, hidden_size)
    hidden_with_time_axis = tf.expand_dims(hidden, 1)

    # score shape == (batch_size, 64, hidden_size)
    score = tf.nn.tanh(self.W1(features) + self.W2(hidden_with_time_axis))

    # attention_weights shape == (batch_size, 64, 1)
    # you get 1 at the last axis because you are applying score to self.V
    attention_weights = tf.nn.softmax(self.V(score), axis=1)

    # context_vector shape after sum == (batch_size, hidden_size)
    context_vector = attention_weights * features
    context_vector = tf.reduce_sum(context_vector, axis=1)

    return context_vector, attention_weights

class CNN_Encoder(tf.keras.Model):
    # Since you have already extracted the features and dumped it using pickle
    # This encoder passes those features through a Fully connected layer
    def __init__(self, embedding_dim):
        super(CNN_Encoder, self).__init__()
        # shape after fc == (batch_size, 64, embedding_dim)
        self.fc = tf.keras.layers.Dense(embedding_dim)

    def call(self, x):
        x = self.fc(x)
        x = tf.nn.relu(x)
        return x

class RNN_Decoder(tf.keras.Model):
  def __init__(self, embedding_dim, units, vocab_size):
    super(RNN_Decoder, self).__init__()
    self.units = units

    self.embedding = tf.keras.layers.Embedding(vocab_size, embedding_dim)
    self.gru = tf.keras.layers.GRU(self.units,
                                   return_sequences=True,
                                   return_state=True,
                                   recurrent_initializer='glorot_uniform')
    self.fc1 = tf.keras.layers.Dense(self.units)
    self.fc2 = tf.keras.layers.Dense(vocab_size)

    self.attention = BahdanauAttention(self.units)

  def call(self, x, features, hidden):
    # defining attention as a separate model
    context_vector, attention_weights = self.attention(features, hidden)

    # x shape after passing through embedding == (batch_size, 1, embedding_dim)
    x = self.embedding(x)

    # x shape after concatenation == (batch_size, 1, embedding_dim + hidden_size)
    x = tf.concat([tf.expand_dims(context_vector, 1), x], axis=-1)

    # passing the concatenated vector to the GRU
    output, state = self.gru(x)

    # shape == (batch_size, max_length, hidden_size)
    x = self.fc1(output)

    # x shape == (batch_size * max_length, hidden_size)
    x = tf.reshape(x, (-1, x.shape[2]))

    # output shape == (batch_size * max_length, vocab)
    x = self.fc2(x)

    return x, state, attention_weights

  def reset_state(self, batch_size):
    return tf.zeros((batch_size, self.units))

encoder = CNN_Encoder(embedding_dim)
decoder = RNN_Decoder(embedding_dim, units, vocab_size)

optimizer = tf.keras.optimizers.Adam()
loss_object = tf.keras.losses.SparseCategoricalCrossentropy(
    from_logits=True, reduction='none')

def loss_function(real, pred):
  mask = tf.math.logical_not(tf.math.equal(real, 0))
  loss_ = loss_object(real, pred)

  mask = tf.cast(mask, dtype=loss_.dtype)
  loss_ *= mask

  return tf.reduce_mean(loss_)

Checkpoint

checkpoint_path = "./checkpoints/train"
ckpt = tf.train.Checkpoint(encoder=encoder,
                           decoder=decoder,
                           optimizer = optimizer)
ckpt_manager = tf.train.CheckpointManager(ckpt, checkpoint_path, max_to_keep=5)

start_epoch = 0
if ckpt_manager.latest_checkpoint:
  start_epoch = int(ckpt_manager.latest_checkpoint.split('-')[-1])

Training

• You extract the features stored in the respective .npy files and then pass those features through the encoder.
• The encoder output, hidden state(initialized to 0) and the decoder input (which is the start token) is passed to the decoder.
• The decoder returns the predictions and the decoder hidden state.
• The decoder hidden state is then passed back into the model and the predictions are used to calculate the loss.
• Use teacher forcing to decide the next input to the decoder.
• Teacher forcing is the technique where the target word is passed as the next input to the decoder.
• The final step is to calculate the gradients and apply it to the optimizer and backpropagate.

# adding this in a separate cell because if you run the training cell
# many times, the loss_plot array will be reset
loss_plot = []

@tf.function
def train_step(img_tensor, target):
  loss = 0

  # initializing the hidden state for each batch
  # because the captions are not related from image to image
  hidden = decoder.reset_state(batch_size=target.shape[0])

  dec_input = tf.expand_dims([tokenizer.word_index['<start>']] * target.shape[0], 1)

  with tf.GradientTape() as tape:
      features = encoder(img_tensor)

      for i in range(1, target.shape[1]):
          # passing the features through the decoder
          predictions, hidden, _ = decoder(dec_input, features, hidden)

          loss += loss_function(target[:, i], predictions)

          # using teacher forcing
          dec_input = tf.expand_dims(target[:, i], 1)

  total_loss = (loss / int(target.shape[1]))

  trainable_variables = encoder.trainable_variables + decoder.trainable_variables

  gradients = tape.gradient(loss, trainable_variables)

  optimizer.apply_gradients(zip(gradients, trainable_variables))

  return loss, total_loss

EPOCHS = 20

for epoch in range(start_epoch, EPOCHS):
    start = time.time()
    total_loss = 0

    for (batch, (img_tensor, target)) in enumerate(dataset):
        batch_loss, t_loss = train_step(img_tensor, target)
        total_loss += t_loss

        if batch % 100 == 0:
            print ('Epoch {} Batch {} Loss {:.4f}'.format(
              epoch + 1, batch, batch_loss.numpy() / int(target.shape[1])))
    # storing the epoch end loss value to plot later
    loss_plot.append(total_loss / num_steps)

    if epoch % 5 == 0:
      ckpt_manager.save()

    print ('Epoch {} Loss {:.6f}'.format(epoch + 1,
                                         total_loss/num_steps))
    print ('Time taken for 1 epoch {} sec\n'.format(time.time() - start))

Epoch 1 Batch 0 Loss 2.1641
Epoch 1 Batch 100 Loss 1.1760
Epoch 1 Batch 200 Loss 1.0655
Epoch 1 Batch 300 Loss 0.9091
Epoch 1 Loss 1.085079
Time taken for 1 epoch 120.04596447944641 sec

Epoch 2 Batch 0 Loss 0.8282
Epoch 2 Batch 100 Loss 0.8594
Epoch 2 Batch 200 Loss 0.8003
Epoch 2 Batch 300 Loss 0.7711
Epoch 2 Loss 0.814150
Time taken for 1 epoch 51.84908151626587 sec

Epoch 3 Batch 0 Loss 0.7733
Epoch 3 Batch 100 Loss 0.8939
Epoch 3 Batch 200 Loss 0.7514
Epoch 3 Batch 300 Loss 0.7305
Epoch 3 Loss 0.735903
Time taken for 1 epoch 51.83467745780945 sec

Epoch 4 Batch 0 Loss 0.7428
Epoch 4 Batch 100 Loss 0.7748
Epoch 4 Batch 200 Loss 0.7225
Epoch 4 Batch 300 Loss 0.7492
Epoch 4 Loss 0.688446
Time taken for 1 epoch 51.909361839294434 sec

Epoch 5 Batch 0 Loss 0.7036
Epoch 5 Batch 100 Loss 0.6540
Epoch 5 Batch 200 Loss 0.6554
Epoch 5 Batch 300 Loss 0.6292
Epoch 5 Loss 0.650351
Time taken for 1 epoch 52.05428624153137 sec

Epoch 6 Batch 0 Loss 0.6319
Epoch 6 Batch 100 Loss 0.6352
Epoch 6 Batch 200 Loss 0.5756
Epoch 6 Batch 300 Loss 0.6273
Epoch 6 Loss 0.616435
Time taken for 1 epoch 52.29492521286011 sec

Epoch 7 Batch 0 Loss 0.5761
Epoch 7 Batch 100 Loss 0.5931
Epoch 7 Batch 200 Loss 0.5976
Epoch 7 Batch 300 Loss 0.5833
Epoch 7 Loss 0.585671
Time taken for 1 epoch 52.13192534446716 sec

Epoch 8 Batch 0 Loss 0.5342
Epoch 8 Batch 100 Loss 0.5098
Epoch 8 Batch 200 Loss 0.5168
Epoch 8 Batch 300 Loss 0.5163
Epoch 8 Loss 0.553834
Time taken for 1 epoch 52.0756254196167 sec

Epoch 9 Batch 0 Loss 0.5078
Epoch 9 Batch 100 Loss 0.5242
Epoch 9 Batch 200 Loss 0.4885
Epoch 9 Batch 300 Loss 0.5613
Epoch 9 Loss 0.524567
Time taken for 1 epoch 52.11587595939636 sec

Epoch 10 Batch 0 Loss 0.5245
Epoch 10 Batch 100 Loss 0.4967
Epoch 10 Batch 200 Loss 0.4655
Epoch 10 Batch 300 Loss 0.4935
Epoch 10 Loss 0.493873
Time taken for 1 epoch 51.99848747253418 sec

Epoch 11 Batch 0 Loss 0.4759
Epoch 11 Batch 100 Loss 0.4637
Epoch 11 Batch 200 Loss 0.4303
Epoch 11 Batch 300 Loss 0.4456
Epoch 11 Loss 0.464059
Time taken for 1 epoch 52.14967370033264 sec

Epoch 12 Batch 0 Loss 0.4271
Epoch 12 Batch 100 Loss 0.4675
Epoch 12 Batch 200 Loss 0.3673
Epoch 12 Batch 300 Loss 0.4140
Epoch 12 Loss 0.434737
Time taken for 1 epoch 51.999250411987305 sec

Epoch 13 Batch 0 Loss 0.4442
Epoch 13 Batch 100 Loss 0.4010
Epoch 13 Batch 200 Loss 0.3925
Epoch 13 Batch 300 Loss 0.4209
Epoch 13 Loss 0.407342
Time taken for 1 epoch 51.98107647895813 sec

Epoch 14 Batch 0 Loss 0.3860
Epoch 14 Batch 100 Loss 0.3648
Epoch 14 Batch 200 Loss 0.3829
Epoch 14 Batch 300 Loss 0.3649
Epoch 14 Loss 0.380663
Time taken for 1 epoch 51.98867869377136 sec

Epoch 15 Batch 0 Loss 0.4159
Epoch 15 Batch 100 Loss 0.3212
Epoch 15 Batch 200 Loss 0.3454
Epoch 15 Batch 300 Loss 0.3550
Epoch 15 Loss 0.354879
Time taken for 1 epoch 52.12524914741516 sec

Epoch 16 Batch 0 Loss 0.3108
Epoch 16 Batch 100 Loss 0.3196
Epoch 16 Batch 200 Loss 0.2848
Epoch 16 Batch 300 Loss 0.3415
Epoch 16 Loss 0.333778
Time taken for 1 epoch 52.17926502227783 sec

Epoch 17 Batch 0 Loss 0.3377
Epoch 17 Batch 100 Loss 0.3221
Epoch 17 Batch 200 Loss 0.3216
Epoch 17 Batch 300 Loss 0.2901
Epoch 17 Loss 0.312425
Time taken for 1 epoch 52.19093203544617 sec

Epoch 18 Batch 0 Loss 0.3126
Epoch 18 Batch 100 Loss 0.3082
Epoch 18 Batch 200 Loss 0.2854
Epoch 18 Batch 300 Loss 0.2873
Epoch 18 Loss 0.289537
Time taken for 1 epoch 52.18620252609253 sec

Epoch 19 Batch 0 Loss 0.3103
Epoch 19 Batch 100 Loss 0.2924
Epoch 19 Batch 200 Loss 0.2763
Epoch 19 Batch 300 Loss 0.2955
Epoch 19 Loss 0.271896
Time taken for 1 epoch 52.26207423210144 sec

Epoch 20 Batch 0 Loss 0.2885
Epoch 20 Batch 100 Loss 0.2319
Epoch 20 Batch 200 Loss 0.2464
Epoch 20 Batch 300 Loss 0.2514
Epoch 20 Loss 0.254451
Time taken for 1 epoch 52.17351746559143 sec

plt.plot(loss_plot)
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.title('Loss Plot')
plt.show()

Caption!

• The evaluate function is similar to the training loop, except you don't use teacher forcing here. The input to the decoder at each time step is its previous predictions along with the hidden state and the encoder output.
• Stop predicting when the model predicts the end token.
• And store the attention weights for every time step.

def evaluate(image):
    attention_plot = np.zeros((max_length, attention_features_shape))

    hidden = decoder.reset_state(batch_size=1)

    temp_input = tf.expand_dims(load_image(image)[0], 0)
    img_tensor_val = image_features_extract_model(temp_input)
    img_tensor_val = tf.reshape(img_tensor_val, (img_tensor_val.shape[0], -1, img_tensor_val.shape[3]))

    features = encoder(img_tensor_val)

    dec_input = tf.expand_dims([tokenizer.word_index['<start>']], 0)
    result = []

    for i in range(max_length):
        predictions, hidden, attention_weights = decoder(dec_input, features, hidden)

        attention_plot[i] = tf.reshape(attention_weights, (-1, )).numpy()

        predicted_id = tf.random.categorical(predictions, 1)[0][0].numpy()
        result.append(tokenizer.index_word[predicted_id])

        if tokenizer.index_word[predicted_id] == '<end>':
            return result, attention_plot

        dec_input = tf.expand_dims([predicted_id], 0)

    attention_plot = attention_plot[:len(result), :]
    return result, attention_plot

def plot_attention(image, result, attention_plot):
    temp_image = np.array(Image.open(image))

    fig = plt.figure(figsize=(10, 10))

    len_result = len(result)
    for l in range(len_result):
        temp_att = np.resize(attention_plot[l], (8, 8))
        ax = fig.add_subplot(len_result//2, len_result//2, l+1)
        ax.set_title(result[l])
        img = ax.imshow(temp_image)
        ax.imshow(temp_att, cmap='gray', alpha=0.6, extent=img.get_extent())

    plt.tight_layout()
    plt.show()

# captions on the validation set
rid = np.random.randint(0, len(img_name_val))
image = img_name_val[rid]
real_caption = ' '.join([tokenizer.index_word[i] for i in cap_val[rid] if i not in [0]])
result, attention_plot = evaluate(image)

print ('Real Caption:', real_caption)
print ('Prediction Caption:', ' '.join(result))
plot_attention(image, result, attention_plot)

Real Caption: <start> a window display of some assorted travel bags <end>
Prediction Caption: a wedding cake that looks like stacked luggage <end>

Try it on your own images

For fun, below we've provided a method you can use to caption your own images with the model we've just trained. Keep in mind, it was trained on a relatively small amount of data, and your images may be different from the training data (so be prepared for weird results!)

image_url = 'https://tensorflow.org/images/surf.jpg'
image_extension = image_url[-4:]
image_path = tf.keras.utils.get_file('image'+image_extension,
                                     origin=image_url)

result, attention_plot = evaluate(image_path)
print ('Prediction Caption:', ' '.join(result))
plot_attention(image_path, result, attention_plot)
# opening the image
Image.open(image_path)

Downloading data from https://tensorflow.org/images/surf.jpg
65536/64400 [==============================] - 0s 3us/step
Prediction Caption: a man that is riding a surfboard in the water <end>

Next steps

Congrats! You've just trained an image captioning model with attention. Next, take a look at this example Neural Machine Translation with Attention. It uses a similar architecture to translate between Spanish and English sentences. You can also experiment with training the code in this notebook on a different dataset.