vgg.py

#! /usr/bin/python
# -*- coding: utf-8 -*-
"""
VGG for ImageNet.

Introduction
----------------
VGG is a convolutional neural network model proposed by K. Simonyan and A. Zisserman
from the University of Oxford in the paper "Very Deep Convolutional Networks for
Large-Scale Image Recognition"  . The model achieves 92.7% top-5 test accuracy in ImageNet,
which is a dataset of over 14 million images belonging to 1000 classes.

Download Pre-trained Model
----------------------------
- Model weights in this example - vgg16_weights.npz : http://www.cs.toronto.edu/~frossard/post/vgg16/
- Model weights in this example - vgg19.npy : https://media.githubusercontent.com/media/tensorlayer/pretrained-models/master/models/
- Caffe VGG 16 model : https://gist.github.com/ksimonyan/211839e770f7b538e2d8#file-readme-md
- Tool to convert the Caffe models to TensorFlow's : https://github.com/ethereon/caffe-tensorflow

Note
------
- For simplified CNN layer see "Convolutional layer (Simplified)"
in read the docs website.
- When feeding other images to the model be sure to properly resize or crop them
beforehand. Distorted images might end up being misclassified. One way of safely
feeding images of multiple sizes is by doing center cropping.

"""

import os

import numpy as np

import tensorlayerx as tlx
from tensorlayerx import logging
from tensorlayerx.files import assign_weights, maybe_download_and_extract
from tensorlayerx.nn import (BatchNorm, Conv2d, Linear, Flatten, Input, Sequential, MaxPool2d)
from tensorlayerx.nn import Module

__all__ = [
    'VGG',
    'vgg16',
    'vgg19',
    'VGG16',
    'VGG19',
    #    'vgg11', 'vgg11_bn', 'vgg13', 'vgg13_bn', 'vgg16', 'vgg16_bn',
    #    'vgg19_bn', 'vgg19',
]

layer_names = [
    ['conv1_1', 'conv1_2'], 'pool1', ['conv2_1', 'conv2_2'], 'pool2',
    ['conv3_1', 'conv3_2', 'conv3_3', 'conv3_4'], 'pool3', ['conv4_1', 'conv4_2', 'conv4_3', 'conv4_4'], 'pool4',
    ['conv5_1', 'conv5_2', 'conv5_3', 'conv5_4'], 'pool5', 'flatten', 'fc1_relu', 'fc2_relu', 'outputs'
]

cfg = {
    'A': [[64], 'M', [128], 'M', [256, 256], 'M', [512, 512], 'M', [512, 512], 'M', 'F', 'fc1', 'fc2', 'O'],
    'B': [[64, 64], 'M', [128, 128], 'M', [256, 256], 'M', [512, 512], 'M', [512, 512], 'M', 'F', 'fc1', 'fc2', 'O'],
    'D':
        [
            [64, 64], 'M', [128, 128], 'M', [256, 256, 256], 'M', [512, 512, 512], 'M', [512, 512, 512], 'M', 'F',
            'fc1', 'fc2', 'O'
        ],
    'E':
        [
            [64, 64], 'M', [128, 128], 'M', [256, 256, 256, 256], 'M', [512, 512, 512, 512], 'M', [512, 512, 512, 512],
            'M', 'F', 'fc1', 'fc2', 'O'
        ],
}

mapped_cfg = {
    'vgg11': 'A',
    'vgg11_bn': 'A',
    'vgg13': 'B',
    'vgg13_bn': 'B',
    'vgg16': 'D',
    'vgg16_bn': 'D',
    'vgg19': 'E',
    'vgg19_bn': 'E'
}

model_urls = {
    'vgg16': 'https://git.openi.org.cn/attachments/760835b9-db71-4a00-8edd-d5ece4b6b522?type=0',
    'vgg19': 'https://git.openi.org.cn/attachments/503c8a6c-705f-4fb6-ba18-03d72b6a949a?type=0'
}

model_saved_name = {'vgg16': 'vgg16_weights.npz', 'vgg19': 'vgg19.npy'}


class VGG(Module):

    def __init__(self, layer_type, batch_norm=False, end_with='outputs', name=None):
        super(VGG, self).__init__(name=name)
        self.end_with = end_with

        config = cfg[mapped_cfg[layer_type]]
        self.make_layer = make_layers(config, batch_norm, end_with)

    def forward(self, inputs):
        """
        inputs : tensor
            Shape [None, 224, 224, 3], value range [0, 1].
        """

#         inputs = inputs * 255 - np.array([123.68, 116.779, 103.939], dtype=np.float32).reshape([1, 1, 1, 3])
        inputs = inputs * 255. - tlx.convert_to_tensor(np.array([123.68, 116.779, 103.939], dtype=np.float32).reshape(-1,1,1))
        out = self.make_layer(inputs)
        return out


def make_layers(config, batch_norm=False, end_with='outputs'):
    layer_list = []
    is_end = False
    for layer_group_idx, layer_group in enumerate(config):
        if isinstance(layer_group, list):
            for idx, layer in enumerate(layer_group):
                layer_name = layer_names[layer_group_idx][idx]
                n_filter = layer
                if idx == 0:
                    if layer_group_idx > 0:
                        in_channels = config[layer_group_idx - 2][-1]
                    else:
                        in_channels = 3
                else:
                    in_channels = layer_group[idx - 1]
                layer_list.append(
                    Conv2d(
                        out_channels=n_filter, kernel_size=(3, 3), stride=(1, 1), act=tlx.ReLU, padding='SAME',
                        in_channels=in_channels, name=layer_name, data_format='channels_first'
                    )
                )
                if batch_norm:
                    layer_list.append(BatchNorm(num_features=n_filter, data_format='channels_first'))
                if layer_name == end_with:
                    is_end = True
                    break
        else:
            layer_name = layer_names[layer_group_idx]
            if layer_group == 'M':
                layer_list.append(MaxPool2d(kernel_size=(2, 2), stride=(2, 2), padding='SAME', name=layer_name, data_format='channels_first'))
            elif layer_group == 'O':
                layer_list.append(Linear(out_features=1000, in_features=4096, name=layer_name))
            elif layer_group == 'F':
                layer_list.append(Flatten(name='flatten'))
            elif layer_group == 'fc1':
                layer_list.append(Linear(out_features=4096, act=tlx.ReLU, in_features=512 * 7 * 7, name=layer_name))
            elif layer_group == 'fc2':
                layer_list.append(Linear(out_features=4096, act=tlx.ReLU, in_features=4096, name=layer_name))
            if layer_name == end_with:
                is_end = True
        if is_end:
            break
    return Sequential(layer_list)

def restore_model(model, layer_type):
    logging.info("Restore pre-trained weights")
    # download weights
    maybe_download_and_extract(model_saved_name[layer_type], 'model', model_urls[layer_type])
    weights = []
    if layer_type == 'vgg16':
        npz = np.load(os.path.join('model', model_saved_name[layer_type]), allow_pickle=True)
        # get weight list
        for val in sorted(npz.items()):
            logging.info("  Loading weights %s in %s" % (str(val[1].shape), val[0]))
            weights.append(val[1])
            if len(model.all_weights) == len(weights):
                break
    elif layer_type == 'vgg19':
        npz = np.load(os.path.join('model', model_saved_name[layer_type]), allow_pickle=True, encoding='latin1').item()
        # get weight list
        for val in sorted(npz.items()):
            logging.info("  Loading %s in %s" % (str(val[1][0].shape), val[0]))
            logging.info("  Loading %s in %s" % (str(val[1][1].shape), val[0]))
            weights.extend(val[1])
            if len(model.all_weights) == len(weights):
                break
    # assign weight values
    if tlx.BACKEND != 'tensorflow':
        for i in range(len(weights)):
            if len(weights[i].shape) == 4:
                weights[i] = np.transpose(weights[i], axes=[3, 2, 0, 1])
    assign_weights(weights, model)
    del weights

def vgg16(pretrained=False, end_with='outputs', mode='dynamic', name=None):
    """Pre-trained VGG16 model.

    Parameters
    ------------
    pretrained : boolean
        Whether to load pretrained weights. Default False.
    end_with : str
        The end point of the model. Default ``fc3_relu`` i.e. the whole model.
    mode : str.
        Model building mode, 'dynamic' or 'static'. Default 'dynamic'.
    name : None or str
        A unique layer name.

    Examples
    ---------
    Classify ImageNet classes with VGG16, see `tutorial_models_vgg.py <https://github.com/tensorlayer/tensorlayer/blob/master/example/tutorial_models_vgg.py>`__
    With TensorLayer
    TODO Modify the usage example according to the model storage location

    >>> # get the whole model, without pre-trained VGG parameters
    >>> vgg = vgg16()
    >>> # get the whole model, restore pre-trained VGG parameters
    >>> vgg = vgg16(pretrained=True)
    >>> # use for inferencing
    >>> output = vgg(img)
    >>> probs = tlx.ops.softmax(output)[0].numpy()

    """

    if mode == 'dynamic':
        model = VGG(layer_type='vgg16', batch_norm=False, end_with=end_with, name=name)
    elif mode == 'static':
        raise NotImplementedError
    else:
        raise Exception("No such mode %s" % mode)
    if pretrained:
        restore_model(model, layer_type='vgg16')
    return model


def vgg19(pretrained=False, end_with='outputs', mode='dynamic', name=None):
    """Pre-trained VGG19 model.

    Parameters
    ------------
    pretrained : boolean
        Whether to load pretrained weights. Default False.
    end_with : str
        The end point of the model. Default ``fc3_relu`` i.e. the whole model.
    mode : str.
        Model building mode, 'dynamic' or 'static'. Default 'dynamic'.
    name : None or str
        A unique layer name.

    Examples
    ---------
    Classify ImageNet classes with VGG19, see `tutorial_models_vgg.py <https://github.com/tensorlayer/TensorLayerX/blob/main/examples/model_zoo/vgg.py>`__
    With TensorLayerx

    >>> # get the whole model, without pre-trained VGG parameters
    >>> vgg = vgg19()
    >>> # get the whole model, restore pre-trained VGG parameters
    >>> vgg = vgg19(pretrained=True)
    >>> # use for inferencing
    >>> output = vgg(img)
    >>> probs = tlx.ops.softmax(output)[0].numpy()

    """
    if mode == 'dynamic':
        model = VGG(layer_type='vgg19', batch_norm=False, end_with=end_with, name=name)
    elif mode == 'static':
        raise NotImplementedError
    else:
        raise Exception("No such mode %s" % mode)
    if pretrained:
        restore_model(model, layer_type='vgg19')
    return model


VGG16 = vgg16
VGG19 = vgg19