add adversarial sample

adversarial/advbox/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) 2017 PaddlePaddle Authors. All Rights Reserved
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+   A set of tools for generating adversarial example on paddle platform 
+"""

adversarial/advbox/attacks/base.py

@@ -0,0 +1,42 @@
+"""
+The base model of the model.
+"""
+from abc import ABCMeta
+#from advbox.base import Model
+import abc
+
+abstractmethod = abc.abstractmethod
+
+class Attack(object):
+    """
+    Abstract base class for adversarial attacks. `Attack` represent an adversarial attack
+    which search an adversarial example. subclass should implement the _apply() method.
+
+    Args:
+        model(Model): an instance of the class advbox.base.Model.
+
+    """
+    __metaclass__ = ABCMeta
+
+    def __init__(self, model):
+        self.model = model
+
+    def __call__(self, image_batch):
+        """
+        Generate the adversarial sample.
+
+        Args:
+        image_batch(list): The image and label tuple list.
+        """
+        adv_img = self._apply(image_batch)
+        return adv_img
+
+    @abstractmethod
+    def _apply(self, image_batch):
+        """
+        Search an adversarial example.
+
+        Args:
+        image_batch(list): The image and label tuple list.
+        """
+        raise NotImplementedError

adversarial/advbox/attacks/gradientsign.py

@@ -0,0 +1,36 @@
+"""
+This module provide the attack method for FGSM's implement.
+"""
+from __future__ import division
+import numpy as np
+from collections import Iterable
+from .base import Attack
+ 
+class GradientSignAttack(Attack):
+    """
+    This attack was originally implemented by Goodfellow et al. (2015) with the
+    infinity norm (and is known as the "Fast Gradient Sign Method"). This is therefore called
+    the Fast Gradient Method.
+    Paper link: https://arxiv.org/abs/1412.6572
+    """
+
+    def _apply(self, image_batch, epsilons=1000):
+        pre_label = np.argmax(self.model.predict(image_batch))
+
+        min_, max_ = self.model.bounds()
+        gradient = self.model.gradient(image_batch)
+        gradient_sign = np.sign(gradient) * (max_ - min_)
+
+        if not isinstance(epsilons, Iterable):
+            epsilons = np.linspace(0, 1, num = epsilons + 1)
+
+        for epsilon in epsilons:
+            adv_img = image_batch[0][0].reshape(gradient_sign.shape) + epsilon * gradient_sign
+            adv_img = np.clip(adv_img, min_, max_)
+            adv_label = np.argmax(self.model.predict([(adv_img, 0)]))
+            #print("pre_label="+str(pre_label)+ " adv_label="+str(adv_label))
+            if pre_label != adv_label:
+                #print(epsilon, pre_label, adv_label)
+                return adv_img
+
+FGSM = GradientSignAttack

adversarial/advbox/models/__init__.py

@@ -0,0 +1,16 @@
+# Copyright (c) 2017 PaddlePaddle Authors. All Rights Reserved
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Paddle model for target of attack 
+"""

adversarial/advbox/models/base.py

@@ -0,0 +1,91 @@
+"""
+The base model of the model.
+"""
+from abc import ABCMeta
+import abc
+
+abstractmethod = abc.abstractmethod
+
+class Model(object):
+
+    """
+    Base class of model to provide attack.
+
+
+    Args:
+        bounds(tuple): The lower and upper bound for the image pixel.
+        channel_axis(int): The index of the axis that represents the color channel.
+        preprocess(tuple): Two element tuple used to preprocess the input. First
+            substract the first element, then divide the second element.
+    """
+    __metaclass__ = ABCMeta
+
+    def __init__(self, bounds, channel_axis, preprocess=None):
+        assert len(bounds) == 2
+        assert channel_axis in [0, 1, 2, 3]
+
+        if preprocess is None:
+            preprocess = (0, 1)
+        self._bounds = bounds
+        self._channel_axis = channel_axis
+        self._preprocess = preprocess
+
+    def bounds(self):
+        """
+        Return the upper and lower bounds of the model.
+        """
+        return self._bounds
+
+    def channel_axis(self):
+        """
+        Return the channel axis of the model.
+        """
+        return self._channel_axis
+
+    def _process_input(self, input_):
+        res = input_
+        sub, div = self._preprocess
+        if sub != 0:
+            res = input_ - sub
+        assert div != 0
+        if div != 1:
+            res /= div
+        return res
+
+    @abstractmethod
+    def predict(self, image_batch):
+        """
+        Calculate the prediction of the image batch.
+
+        Args:
+            image_batch(numpy.ndarray): image batch of shape (batch_size, height, width, channels).
+
+        Return:
+            numpy.ndarray: predictions of the images with shape (batch_size, num_of_classes).
+        """
+        raise NotImplementedError
+
+    @abstractmethod
+    def num_classes(self):
+        """
+        Determine the number of the classes
+
+        Return:
+            int: the number of the classes
+        """
+        raise NotImplementedError
+
+    @abstractmethod
+    def gradient(self, image_batch):
+        """
+        Calculate the gradient of the cross-entropy loss w.r.t the image.
+
+        Args:
+            image(numpy.ndarray): image with shape (height, width, channel)
+            label(int): image label used to cal gradient.
+
+        Return:
+            numpy.ndarray: gradient of the cross-entropy loss w.r.t the image with
+                the shape (height, width, channel).
+        """
+        raise NotImplementedError

adversarial/advbox/models/paddle.py

@@ -0,0 +1,106 @@
+from __future__ import absolute_import
+
+import numpy as np
+import paddle.v2 as paddle
+import paddle.v2.fluid as fluid
+from paddle.v2.fluid.framework import program_guard
+
+from .base import Model
+
+class PaddleModel(Model):
+    """
+    Create a PaddleModel instance.
+    When you need to generate a adversarial sample, you should construct an instance of PaddleModel.
+
+    Args:
+        program(paddle.v2.fluid.framework.Program): The program of the model which generate the adversarial sample.
+        input_name(string): The name of the input.
+        logits_name(string): The name of the logits.
+        predict_name(string): The name of the predict.
+        cost_name(string): The name of the loss in the program.
+    """
+
+    def __init__(self,
+                 program,
+                 input_name,
+                 logits_name,
+                 predict_name,
+                 cost_name,
+                 bounds,
+                 channel_axis=3,
+                 preprocess=None):
+        super(PaddleModel, self).__init__(
+            bounds=bounds,
+            channel_axis=channel_axis,
+            preprocess=preprocess)
+
+        if preprocess is None:
+            preprocess = (0, 1)
+
+        self._program = program
+        self._place = fluid.CPUPlace()
+        self._exe = fluid.Executor(self._place)
+
+        self._input_name = input_name
+        self._logits_name = logits_name
+        self._predict_name = predict_name
+        self._cost_name = cost_name
+
+        # gradient
+        loss = self._program.block(0).var(self._cost_name)
+        param_grads = fluid.backward.append_backward(loss, parameter_list=[self._input_name])
+        self._gradient = param_grads[0][1]
+
+    def predict(self, image_batch):
+        """
+            Predict the label of the image_batch.
+
+            Args:
+                image_batch(list): The image and label tuple list.
+            Return:
+                numpy.ndarray: predictions of the images with shape (batch_size, num_of_classes).
+        """
+        feeder = fluid.DataFeeder(
+            feed_list=[self._input_name, self._logits_name], 
+            place=self._place, 
+            program=self._program
+        )
+        predict_var = self._program.block(0).var(self._predict_name)
+        predict = self._exe.run(
+            self._program,
+            feed=feeder.feed(image_batch),
+            fetch_list=[predict_var]
+        )
+        return predict
+
+    def num_classes(self):
+        """
+            Calculate the number of classes of the output label. 
+
+        Return:
+            int: the number of classes
+        """
+        predict_var = self._program.block(0).var(self._predict_name)
+        assert len(predict_var.shape) == 2
+        return predict_var.shape[1]
+
+    def gradient(self, image_batch):
+        """
+        Calculate the gradient of the loss w.r.t the input.
+
+        Args:
+            image_batch(list): The image and label tuple list.
+        Return:
+            list: The list of the gradient of the image.
+        """
+        feeder = fluid.DataFeeder(
+            feed_list=[self._input_name, self._logits_name],
+            place=self._place, 
+            program=self._program
+        )
+
+        grad, = self._exe.run(
+            self._program,
+            feed=feeder.feed(image_batch),
+            fetch_list=[self._gradient])
+        return grad

adversarial/advbox/tutorials/tutorial_model.py

@@ -0,0 +1,32 @@
+################################################################################
+#
+# Copyright (c) 2017 Baidu.com, Inc. All Rights Reserved
+#
+################################################################################
+"""
+
+A pure Paddlepaddle implementation of a neural network.
+
+"""
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+from __future__ import unicode_literals
+
+import paddle.v2 as paddle
+import paddle.v2.fluid as fluid
+from advbox import Model
+
+def main():
+    """
+	example main function
+    """
+    model_dir = "./mnist_model"
+    place = fluid.CPUPlace()
+    exe = fluid.Executor(place)
+    program, feed_var_names, fetch_vars = fluid.io.load_inferfence_model(model_dir, exe)
+    print(program)
+
+if __name__ == "__main__":
+    main()

adversarial/fluid_mnist.py

@@ -0,0 +1,91 @@
+"""
+CNN on mnist data using fluid api of paddlepaddle
+"""
+import paddle.v2 as paddle
+import paddle.v2.fluid as fluid
+
+def mnist_cnn_model(img):
+    """
+    Mnist cnn model
+
+    Args:
+        img(Varaible): the input image to be recognized
+
+    Returns:
+        Variable: the label prediction
+    """
+    #conv1 = fluid.nets.conv2d()
+    conv_pool_1 = fluid.nets.simple_img_conv_pool(
+        input=img,
+        num_filters=20,
+        filter_size=5,
+        pool_size=2,
+        pool_stride=2,
+        act='relu')
+
+    conv_pool_2 = fluid.nets.simple_img_conv_pool(
+        input=conv_pool_1,
+        num_filters=50,
+        filter_size=5,
+        pool_size=2,
+        pool_stride=2,
+        act='relu')
+
+    logits = fluid.layers.fc(
+        input=conv_pool_2,
+        size=10,
+        act='softmax')
+    return logits
+
+
+def main():
+    """
+    Train the cnn model on mnist datasets
+    """
+    img = fluid.layers.data(name='img', shape=[1, 28, 28], dtype='float32')
+    label = fluid.layers.data(name='label', shape=[1], dtype='int64')
+    logits = mnist_cnn_model(img)
+    cost = fluid.layers.cross_entropy(input=logits, label=label)
+    avg_cost = fluid.layers.mean(x=cost)
+    optimizer = fluid.optimizer.Adam(learning_rate=0.01)
+    optimizer.minimize(avg_cost)
+
+    accuracy = fluid.evaluator.Accuracy(input=logits, label=label)
+
+    BATCH_SIZE = 50
+    PASS_NUM = 3
+    ACC_THRESHOLD = 0.98
+    LOSS_THRESHOLD = 10.0
+    train_reader = paddle.batch(
+        paddle.reader.shuffle(
+            paddle.dataset.mnist.train(), buf_size=500),
+        batch_size=BATCH_SIZE)
+
+    place = fluid.CPUPlace()
+    exe = fluid.Executor(place)
+    feeder = fluid.DataFeeder(feed_list=[img, label], place=place)
+    exe.run(fluid.default_startup_program())
+
+    for pass_id in range(PASS_NUM):
+        accuracy.reset(exe)
+        for data in train_reader():
+            loss, acc = exe.run(fluid.default_main_program(),
+                                feed=feeder.feed(data),
+                                fetch_list=[avg_cost] + accuracy.metrics)
+            pass_acc = accuracy.eval(exe)
+            print("pass_id=" + str(pass_id) + " acc=" + str(acc) + " pass_acc=" +
+                  str(pass_acc))
+            # print loss, acc
+            if loss < LOSS_THRESHOLD and pass_acc > ACC_THRESHOLD:
+                # if avg cost less than 10.0 and accuracy is larger than 0.9, we think our code is good.
+                break
+#                exit(0)
+
+        pass_acc = accuracy.eval(exe)
+        print("pass_id=" + str(pass_id) + " pass_acc=" + str(pass_acc))
+    fluid.io.save_params(exe, dirname='./mnist', main_program=fluid.default_main_program())
+    print('train mnist done')
+    exit(1)
+
+if __name__ == '__main__':
+    main()

adversarial/mnist_fgsm.py

@@ -0,0 +1,113 @@
+"""
+This attack was originally implemented by Goodfellow et al. (2015) with the
+infinity norm (and is known as the "Fast Gradient Sign Method"). This is therefore called
+the Fast Gradient Method.
+Paper link: https://arxiv.org/abs/1412.6572
+"""
+
+import numpy as np
+import paddle.v2 as paddle
+import paddle.v2.fluid as fluid
+
+BATCH_SIZE = 50
+PASS_NUM = 1
+EPS = 0.3
+CLIP_MIN = -1
+CLIP_MAX = 1
+PASS_NUM = 1
+
+def mnist_cnn_model(img):
+    """
+    Mnist cnn model
+
+    Args:
+        img(Varaible): the input image to be recognized
+
+    Returns:
+        Variable: the label prediction
+    """
+    #conv1 = fluid.nets.conv2d()
+    conv_pool_1 = fluid.nets.simple_img_conv_pool(
+        input=img,
+        num_filters=20,
+        filter_size=5,
+        pool_size=2,
+        pool_stride=2,
+        act='relu')
+
+    conv_pool_2 = fluid.nets.simple_img_conv_pool(
+        input=conv_pool_1,
+        num_filters=50,
+        filter_size=5,
+        pool_size=2,
+        pool_stride=2,
+        act='relu')
+
+    logits = fluid.layers.fc(
+        input=conv_pool_2,
+        size=10,
+        act='softmax')
+    return logits
+
+
+def main():
+    """
+    Generate adverserial example and evaluate accuracy on mnist using FGSM
+    """
+
+    images = fluid.layers.data(name='pixel', shape=[1, 28, 28], dtype='float32')
+    # The gradient should flow
+    images.stop_gradient = False
+    label = fluid.layers.data(name='label', shape=[1], dtype='int64')
+
+    predict = mnist_cnn_model(images)
+    cost = fluid.layers.cross_entropy(input=predict, label=label)
+    avg_cost = fluid.layers.mean(x=cost)
+
+    # Cal gradient of input
+    params_grads = fluid.backward.append_backward_ops(avg_cost, parameter_list=['pixel'])
+    # data batch
+    train_reader = paddle.batch(
+        paddle.reader.shuffle(
+            paddle.dataset.mnist.train(), buf_size=500),
+        batch_size=BATCH_SIZE)
+
+    accuracy = fluid.evaluator.Accuracy(input=predict, label=label)
+    place = fluid.CPUPlace()
+    exe = fluid.Executor(place)
+    accuracy.reset(exe)
+    #exe.run(fluid.default_startup_program())
+    feeder = fluid.DataFeeder(feed_list=[images, label], place=place)
+    for pass_id in range(PASS_NUM):
+        fluid.io.load_params(exe, "./mnist/", main_program=fluid.default_main_program())
+        for data in train_reader():
+            # cal gradient and eval accuracy
+            ps, acc = exe.run(
+                fluid.default_main_program(),
+                feed=feeder.feed(data),
+                fetch_list=[params_grads[0][1]]+accuracy.metrics)
+            labels = []
+            for idx, _ in enumerate(data):
+                labels.append(data[idx][1])
+            # generate adversarial example
+            batch_num = ps.shape[0]
+            new_data = []
+            for i in range(batch_num):
+                adv_img = np.reshape(data[0][0], (1, 28, 28)) + EPS * np.sign(ps[i])
+                adv_img = np.clip(adv_img, CLIP_MIN, CLIP_MAX)
+                #adv_imgs.append(adv_img)
+                t = (adv_img, data[0][1])
+                new_data.append(t)
+
+            # predict label
+            predict_label, = exe.run(
+                fluid.default_main_program(),
+                feed=feeder.feed(new_data),
+                fetch_list=[predict])
+            adv_labels = np.argmax(predict_label, axis=1)
+            batch_accuracy = np.mean(np.equal(labels, adv_labels))
+            print "pass_id=" + str(pass_id) + " acc=" + str(acc)+ " adv_acc=" + str(batch_accuracy)
+
+
+if __name__ == "__main__":
+    main()

adversarial/mnist_tutorial_fgsm.py

@@ -0,0 +1,94 @@
+"""
+FGSM demos on mnist using advbox tool.
+"""
+import paddle.v2 as paddle
+import paddle.v2.fluid as fluid
+import matplotlib.pyplot as plt
+import numpy as np
+
+from advbox.models.paddle import PaddleModel
+from advbox.attacks.gradientsign import GradientSignAttack
+
+def cnn_model(img):
+    """
+    Mnist cnn model
+    Args:
+        img(Varaible): the input image to be recognized
+    Returns:
+        Variable: the label prediction
+    """
+    #conv1 = fluid.nets.conv2d()
+    conv_pool_1 = fluid.nets.simple_img_conv_pool(
+            input=img,
+            num_filters=20,
+            filter_size=5,
+            pool_size=2,
+            pool_stride=2,
+            act='relu')
+
+    conv_pool_2 = fluid.nets.simple_img_conv_pool(
+            input=conv_pool_1,
+            num_filters=50,
+            filter_size=5,
+            pool_size=2,
+            pool_stride=2,
+            act='relu')
+
+    logits = fluid.layers.fc(
+            input=conv_pool_2,
+            size=10,
+            act='softmax')
+    return logits
+
+
+def main():
+    """
+    Advbox demo which demonstrate how to use advbox.
+    """
+    IMG_NAME = 'img'
+    LABEL_NAME = 'label'
+
+    img = fluid.layers.data(name=IMG_NAME, shape=[1, 28, 28], dtype='float32')
+    # gradient should flow
+    img.stop_gradient = False
+    label = fluid.layers.data(name=LABEL_NAME, shape=[1], dtype='int64')
+    logits = cnn_model(img)
+    cost = fluid.layers.cross_entropy(input=logits, label=label)
+    avg_cost = fluid.layers.mean(x=cost)
+
+    place = fluid.CPUPlace()
+    exe = fluid.Executor(place)
+
+    BATCH_SIZE = 1
+    train_reader = paddle.batch(
+        paddle.reader.shuffle(
+            paddle.dataset.mnist.train(), buf_size=500),
+        batch_size=BATCH_SIZE)
+    feeder = fluid.DataFeeder(
+        feed_list=[IMG_NAME, LABEL_NAME], 
+        place=place, 
+        program=fluid.default_main_program()
+    )
+
+    fluid.io.load_params(exe, "./mnist/", main_program=fluid.default_main_program())
+
+    # advbox demo
+    m = PaddleModel(
+        fluid.default_main_program(), 
+        IMG_NAME,
+        LABEL_NAME, 
+        logits.name, 
+        avg_cost.name, 
+        (-1, 1)
+    )
+    att = GradientSignAttack(m)
+    for data in train_reader():
+        # fgsm attack
+        adv_img = att(data)
+        plt.imshow(n[0][0], cmap='Greys_r')
+        plt.show()
+        #np.save('adv_img', adv_img)
+        break
+    
+if __name__ == '__main__':
+    main()