Image2Image.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
# File: Image2Image.py
# Author: Yuxin Wu <ppwwyyxxc@gmail.com>

import cv2
import numpy as np
import tensorflow as tf
import glob
import os
import argparse


from tensorpack import *
from tensorpack.utils.viz import stack_patches
from tensorpack.tfutils.summary import add_moving_summary
from tensorpack.tfutils.scope_utils import auto_reuse_variable_scope
from GAN import GANTrainer, GANModelDesc

"""
To train Image-to-Image translation model with image pairs:
    ./Image2Image.py --data /path/to/datadir --mode {AtoB,BtoA}
    # datadir should contain jpg images of shpae 2s x s, formed by A and B
    # you can download some data from the original authors:
    # https://people.eecs.berkeley.edu/~tinghuiz/projects/pix2pix/datasets/

Speed:
    On GTX1080 with BATCH=1, the speed is about 9.3it/s (the original torch version is 9.5it/s)

Training visualization will appear be in tensorboard.
To visualize on test set:
    ./Image2Image.py --sample --data /path/to/test/datadir --mode {AtoB,BtoA} --load model

"""

BATCH = 1
IN_CH = 3
OUT_CH = 3
LAMBDA = 100
NF = 64  # number of filter


def BNLReLU(x, name=None):
    x = BatchNorm('bn', x)
    return tf.nn.leaky_relu(x, alpha=0.2, name=name)


def visualize_tensors(name, imgs, scale_func=lambda x: (x + 1.) * 128., max_outputs=1):
    """Generate tensor for TensorBoard (casting, clipping)

    Args:
        name: name for visualization operation
        *imgs: multiple tensors as list
        scale_func: scale input tensors to fit range [0, 255]

    Example:
        visualize_tensors('viz1', [img1])
        visualize_tensors('viz2', [img1, img2, img3], max_outputs=max(30, BATCH))
    """
    xy = scale_func(tf.concat(imgs, axis=2))
    xy = tf.cast(tf.clip_by_value(xy, 0, 255), tf.uint8, name='viz')
    tf.summary.image(name, xy, max_outputs=30)


class Model(GANModelDesc):
    def inputs(self):
        SHAPE = 256
        return [tf.placeholder(tf.float32, (None, SHAPE, SHAPE, IN_CH), 'input'),
                tf.placeholder(tf.float32, (None, SHAPE, SHAPE, OUT_CH), 'output')]

    def generator(self, imgs):
        # imgs: input: 256x256xch
        # U-Net structure, it's slightly different from the original on the location of relu/lrelu
        with argscope(BatchNorm, use_local_stat=True), \
                argscope(Dropout, is_training=True):
            # always use local stat for BN, and apply dropout even in testing
            with argscope(Conv2D, kernel_size=4, strides=2, activation=BNLReLU):
                e1 = Conv2D('conv1', imgs, NF, activation=tf.nn.leaky_relu)
                e2 = Conv2D('conv2', e1, NF * 2)
                e3 = Conv2D('conv3', e2, NF * 4)
                e4 = Conv2D('conv4', e3, NF * 8)
                e5 = Conv2D('conv5', e4, NF * 8)
                e6 = Conv2D('conv6', e5, NF * 8)
                e7 = Conv2D('conv7', e6, NF * 8)
                e8 = Conv2D('conv8', e7, NF * 8, activation=BNReLU)  # 1x1
            with argscope(Conv2DTranspose, activation=BNReLU, kernel_size=4, strides=2):
                return (LinearWrap(e8)
                        .Conv2DTranspose('deconv1', NF * 8)
                        .Dropout()
                        .ConcatWith(e7, 3)
                        .Conv2DTranspose('deconv2', NF * 8)
                        .Dropout()
                        .ConcatWith(e6, 3)
                        .Conv2DTranspose('deconv3', NF * 8)
                        .Dropout()
                        .ConcatWith(e5, 3)
                        .Conv2DTranspose('deconv4', NF * 8)
                        .ConcatWith(e4, 3)
                        .Conv2DTranspose('deconv5', NF * 4)
                        .ConcatWith(e3, 3)
                        .Conv2DTranspose('deconv6', NF * 2)
                        .ConcatWith(e2, 3)
                        .Conv2DTranspose('deconv7', NF * 1)
                        .ConcatWith(e1, 3)
                        .Conv2DTranspose('deconv8', OUT_CH, activation=tf.tanh)())

    @auto_reuse_variable_scope
    def discriminator(self, inputs, outputs):
        """ return a (b, 1) logits"""
        l = tf.concat([inputs, outputs], 3)
        with argscope(Conv2D, kernel_size=4, strides=2, activation=BNLReLU):
            l = (LinearWrap(l)
                 .Conv2D('conv0', NF, activation=tf.nn.leaky_relu)
                 .Conv2D('conv1', NF * 2)
                 .Conv2D('conv2', NF * 4)
                 .Conv2D('conv3', NF * 8, strides=1, padding='VALID')
                 .Conv2D('convlast', 1, strides=1, padding='VALID', activation=tf.identity)())
        return l

    def build_graph(self, input, output):
        input, output = input / 128.0 - 1, output / 128.0 - 1

        with argscope([Conv2D, Conv2DTranspose], kernel_initializer=tf.truncated_normal_initializer(stddev=0.02)):
            with tf.variable_scope('gen'):
                fake_output = self.generator(input)
            with tf.variable_scope('discrim'):
                real_pred = self.discriminator(input, output)
                fake_pred = self.discriminator(input, fake_output)

        self.build_losses(real_pred, fake_pred)
        errL1 = tf.reduce_mean(tf.abs(fake_output - output), name='L1_loss')
        self.g_loss = tf.add(self.g_loss, LAMBDA * errL1, name='total_g_loss')
        add_moving_summary(errL1, self.g_loss)

        # tensorboard visualization
        if IN_CH == 1:
            input = tf.image.grayscale_to_rgb(input)
        if OUT_CH == 1:
            output = tf.image.grayscale_to_rgb(output)
            fake_output = tf.image.grayscale_to_rgb(fake_output)

        visualize_tensors('input,output,fake', [input, output, fake_output], max_outputs=max(30, BATCH))

        self.collect_variables()

    def optimizer(self):
        lr = tf.get_variable('learning_rate', initializer=2e-4, trainable=False)
        return tf.train.AdamOptimizer(lr, beta1=0.5, epsilon=1e-3)


def split_input(img):
    """
    img: an RGB image of shape (s, 2s, 3).
    :return: [input, output]
    """
    # split the image into left + right pairs
    s = img.shape[0]
    assert img.shape[1] == 2 * s
    input, output = img[:, :s, :], img[:, s:, :]
    if args.mode == 'BtoA':
        input, output = output, input
    if IN_CH == 1:
        input = cv2.cvtColor(input, cv2.COLOR_RGB2GRAY)[:, :, np.newaxis]
    if OUT_CH == 1:
        output = cv2.cvtColor(output, cv2.COLOR_RGB2GRAY)[:, :, np.newaxis]
    return [input, output]


def get_data():
    datadir = args.data
    imgs = glob.glob(os.path.join(datadir, '*.jpg'))
    ds = ImageFromFile(imgs, channel=3, shuffle=True)

    ds = MapData(ds, lambda dp: split_input(dp[0]))
    augs = [imgaug.Resize(286), imgaug.RandomCrop(256)]
    ds = AugmentImageComponents(ds, augs, (0, 1))
    ds = BatchData(ds, BATCH)
    ds = PrefetchData(ds, 100, 1)
    return ds


def sample(datadir, model_path):
    pred = PredictConfig(
        session_init=get_model_loader(model_path),
        model=Model(),
        input_names=['input', 'output'],
        output_names=['viz'])

    imgs = glob.glob(os.path.join(datadir, '*.jpg'))
    ds = ImageFromFile(imgs, channel=3, shuffle=True)
    ds = MapData(ds, lambda dp: split_input(dp[0]))
    ds = AugmentImageComponents(ds, [imgaug.Resize(256)], (0, 1))
    ds = BatchData(ds, 6)

    pred = SimpleDatasetPredictor(pred, ds)
    for o in pred.get_result():
        o = o[0][:, :, :, ::-1]
        stack_patches(o, nr_row=3, nr_col=2, viz=True)


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--gpu', help='comma separated list of GPU(s) to use.')
    parser.add_argument('--load', help='load model')
    parser.add_argument('--sample', action='store_true', help='run sampling')
    parser.add_argument('--data', help='Image directory', required=True)
    parser.add_argument('--mode', choices=['AtoB', 'BtoA'], default='AtoB')
    parser.add_argument('-b', '--batch', type=int, default=1)
    global args
    args = parser.parse_args()
    if args.gpu:
        os.environ['CUDA_VISIBLE_DEVICES'] = args.gpu

    BATCH = args.batch

    if args.sample:
        assert args.load
        sample(args.data, args.load)
    else:
        logger.auto_set_dir()

        data = QueueInput(get_data())

        GANTrainer(data, Model()).train_with_defaults(
            callbacks=[
                PeriodicTrigger(ModelSaver(), every_k_epochs=3),
                ScheduledHyperParamSetter('learning_rate', [(200, 1e-4)])
            ],
            steps_per_epoch=data.size(),
            max_epoch=300,
            session_init=SaverRestore(args.load) if args.load else None
        )