dataset.py

import os
import sys
import re
import six
import math
import lmdb
import torch

from augmentation.weather import Fog, Snow, Frost
from augmentation.warp import Curve, Distort, Stretch
from augmentation.geometry import Rotate, Perspective, Shrink, TranslateX, TranslateY
from augmentation.pattern import VGrid, HGrid, Grid, RectGrid, EllipseGrid
from augmentation.noise import GaussianNoise, ShotNoise, ImpulseNoise, SpeckleNoise
from augmentation.blur import GaussianBlur, DefocusBlur, MotionBlur, GlassBlur, ZoomBlur
from augmentation.camera import Contrast, Brightness, JpegCompression, Pixelate
from augmentation.weather import Fog, Snow, Frost, Rain, Shadow
from augmentation.process import Posterize, Solarize, Invert, Equalize, AutoContrast, Sharpness, Color

from natsort import natsorted
from PIL import Image
import PIL.ImageOps
import numpy as np
from torch.utils.data import Dataset, ConcatDataset, Subset
from torch._utils import _accumulate
import torchvision.transforms as transforms
import torchvision.transforms.functional as TF


class Batch_Balanced_Dataset(object):

    def __init__(self, opt):
        """
        Modulate the data ratio in the batch.
        For example, when select_data is "MJ-ST" and batch_ratio is "0.5-0.5",
        the 50% of the batch is filled with MJ and the other 50% of the batch is filled with ST.
        """
        log = open(f'./saved_models/{opt.exp_name}/log_dataset.txt', 'a')
        dashed_line = '-' * 80
        print(dashed_line)
        log.write(dashed_line + '\n')
        print(f'dataset_root: {opt.train_data}\nopt.select_data: {opt.select_data}\nopt.batch_ratio: {opt.batch_ratio}')
        log.write(f'dataset_root: {opt.train_data}\nopt.select_data: {opt.select_data}\nopt.batch_ratio: {opt.batch_ratio}\n')
        assert len(opt.select_data) == len(opt.batch_ratio)

        _AlignCollate = AlignCollate(imgH=opt.imgH, imgW=opt.imgW, keep_ratio_with_pad=opt.PAD, opt=opt)
        self.data_loader_list = []
        self.dataloader_iter_list = []
        batch_size_list = []
        Total_batch_size = 0
        for selected_d, batch_ratio_d in zip(opt.select_data, opt.batch_ratio):
            _batch_size = max(round(opt.batch_size * float(batch_ratio_d)), 1)
            print(dashed_line)
            log.write(dashed_line + '\n')
            _dataset, _dataset_log = hierarchical_dataset(root=opt.train_data, opt=opt, select_data=[selected_d])
            total_number_dataset = len(_dataset)
            log.write(_dataset_log)

            """
            The total number of data can be modified with opt.total_data_usage_ratio.
            ex) opt.total_data_usage_ratio = 1 indicates 100% usage, and 0.2 indicates 20% usage.
            See 4.2 section in our paper.
            """
            number_dataset = int(total_number_dataset * float(opt.total_data_usage_ratio))
            dataset_split = [number_dataset, total_number_dataset - number_dataset]
            indices = range(total_number_dataset)
            _dataset, _ = [Subset(_dataset, indices[offset - length:offset])
                           for offset, length in zip(_accumulate(dataset_split), dataset_split)]
            selected_d_log = f'num total samples of {selected_d}: {total_number_dataset} x {opt.total_data_usage_ratio} (total_data_usage_ratio) = {len(_dataset)}\n'
            selected_d_log += f'num samples of {selected_d} per batch: {opt.batch_size} x {float(batch_ratio_d)} (batch_ratio) = {_batch_size}'
            print(selected_d_log)
            log.write(selected_d_log + '\n')
            batch_size_list.append(str(_batch_size))
            Total_batch_size += _batch_size

            _data_loader = torch.utils.data.DataLoader(
                _dataset, batch_size=_batch_size,
                shuffle=True,
                num_workers=int(opt.workers),
                collate_fn=_AlignCollate, pin_memory=True)
            self.data_loader_list.append(_data_loader)
            self.dataloader_iter_list.append(iter(_data_loader))

        Total_batch_size_log = f'{dashed_line}\n'
        batch_size_sum = '+'.join(batch_size_list)
        Total_batch_size_log += f'Total_batch_size: {batch_size_sum} = {Total_batch_size}\n'
        Total_batch_size_log += f'{dashed_line}'
        opt.batch_size = Total_batch_size

        print(Total_batch_size_log)
        log.write(Total_batch_size_log + '\n')
        log.close()

    def get_batch(self):
        balanced_batch_images = []
        balanced_batch_texts = []

        for i, data_loader_iter in enumerate(self.dataloader_iter_list):
            try:
                image, text = data_loader_iter.next()
                balanced_batch_images.append(image)
                balanced_batch_texts += text
            except StopIteration:
                self.dataloader_iter_list[i] = iter(self.data_loader_list[i])
                image, text = self.dataloader_iter_list[i].next()
                balanced_batch_images.append(image)
                balanced_batch_texts += text
            except ValueError:
                pass

        balanced_batch_images = torch.cat(balanced_batch_images, 0)

        return balanced_batch_images, balanced_batch_texts


def hierarchical_dataset(root, opt, select_data='/'):
    """ select_data='/' contains all sub-directory of root directory """
    dataset_list = []
    dataset_log = f'dataset_root:    {root}\t dataset: {select_data[0]}'
    print(dataset_log)
    dataset_log += '\n'
    for dirpath, dirnames, filenames in os.walk(root+'/'):
        if not dirnames:
            select_flag = False
            for selected_d in select_data:
                if selected_d in dirpath:
                    select_flag = True
                    break

            if select_flag:
                dataset = LmdbDataset(dirpath, opt)
                sub_dataset_log = f'sub-directory:\t/{os.path.relpath(dirpath, root)}\t num samples: {len(dataset)}'
                print(sub_dataset_log)
                dataset_log += f'{sub_dataset_log}\n'
                dataset_list.append(dataset)

    concatenated_dataset = ConcatDataset(dataset_list)

    return concatenated_dataset, dataset_log


class LmdbDataset(Dataset):

    def __init__(self, root, opt):

        self.root = root
        self.opt = opt
        self.env = lmdb.open(root, max_readers=32, readonly=True, lock=False, readahead=False, meminit=False)
        if not self.env:
            print('cannot create lmdb from %s' % (root))
            sys.exit(0)

        with self.env.begin(write=False) as txn:
            nSamples = int(txn.get('num-samples'.encode()))
            self.nSamples = nSamples

            if self.opt.data_filtering_off:
                # for fast check or benchmark evaluation with no filtering
                self.filtered_index_list = [index + 1 for index in range(self.nSamples)]
            else:
                """ Filtering part
                If you want to evaluate IC15-2077 & CUTE datasets which have special character labels,
                use --data_filtering_off and only evaluate on alphabets and digits.
                see https://github.com/clovaai/deep-text-recognition-benchmark/blob/6593928855fb7abb999a99f428b3e4477d4ae356/dataset.py#L190-L192

                And if you want to evaluate them with the model trained with --sensitive option,
                use --sensitive and --data_filtering_off,
                see https://github.com/clovaai/deep-text-recognition-benchmark/blob/dff844874dbe9e0ec8c5a52a7bd08c7f20afe704/test.py#L137-L144
                """
                self.filtered_index_list = []
                for index in range(self.nSamples):
                    index += 1  # lmdb starts with 1
                    label_key = 'label-%09d'.encode() % index
                    label = txn.get(label_key).decode('utf-8')

                    if len(label) > self.opt.batch_max_length:
                        # print(f'The length of the label is longer than max_length: length
                        # {len(label)}, {label} in dataset {self.root}')
                        continue

                    # By default, images containing characters which are not in opt.character are filtered.
                    # You can add [UNK] token to `opt.character` in utils.py instead of this filtering.
                    out_of_char = f'[^{self.opt.character}]'
                    if re.search(out_of_char, label.lower()):
                        continue

                    self.filtered_index_list.append(index)

                self.nSamples = len(self.filtered_index_list)

    def __len__(self):
        return self.nSamples

    def __getitem__(self, index):
        assert index <= len(self), 'index range error'
        index = self.filtered_index_list[index]

        with self.env.begin(write=False) as txn:
            label_key = 'label-%09d'.encode() % index
            label = txn.get(label_key).decode('utf-8')
            img_key = 'image-%09d'.encode() % index
            imgbuf = txn.get(img_key)

            buf = six.BytesIO()
            buf.write(imgbuf)
            buf.seek(0)
            try:
                if self.opt.rgb:
                    img = Image.open(buf).convert('RGB')  # for color image
                else:
                    img = Image.open(buf).convert('L')

            except IOError:
                print(f'Corrupted image for {index}')
                # make dummy image and dummy label for corrupted image.
                if self.opt.rgb:
                    img = Image.new('RGB', (self.opt.imgW, self.opt.imgH))
                else:
                    img = Image.new('L', (self.opt.imgW, self.opt.imgH))
                label = '[dummy_label]'

            if not self.opt.sensitive:
                label = label.lower()

            # We only train and evaluate on alphanumerics (or pre-defined character set in train.py)
            out_of_char = f'[^{self.opt.character}]'
            label = re.sub(out_of_char, '', label)

        return (img, label)


class RawDataset(Dataset):

    def __init__(self, root, opt):
        self.opt = opt
        self.image_path_list = []
        for dirpath, dirnames, filenames in os.walk(root):
            for name in filenames:
                _, ext = os.path.splitext(name)
                ext = ext.lower()
                if ext == '.jpg' or ext == '.jpeg' or ext == '.png':
                    self.image_path_list.append(os.path.join(dirpath, name))

        self.image_path_list = natsorted(self.image_path_list)
        self.nSamples = len(self.image_path_list)

    def __len__(self):
        return self.nSamples

    def __getitem__(self, index):

        try:
            if self.opt.rgb:
                img = Image.open(self.image_path_list[index]).convert('RGB')  # for color image
            else:
                img = Image.open(self.image_path_list[index]).convert('L')

        except IOError:
            print(f'Corrupted image for {index}')
            # make dummy image and dummy label for corrupted image.
            if self.opt.rgb:
                img = Image.new('RGB', (self.opt.imgW, self.opt.imgH))
            else:
                img = Image.new('L', (self.opt.imgW, self.opt.imgH))

        return (img, self.image_path_list[index])


def isless(prob=0.5):
    return np.random.uniform(0,1) < prob

class DataAugment(object):
    '''
    Supports with and without data augmentation 
    '''
    def __init__(self, opt):
        self.opt = opt

        if not opt.eval:
            self.process = [Posterize(), Solarize(), Invert(), Equalize(), AutoContrast(), Sharpness(), Color()]
            self.camera = [Contrast(), Brightness(), JpegCompression(), Pixelate()]

            self.pattern = [VGrid(), HGrid(), Grid(), RectGrid(), EllipseGrid()]

            self.noise = [GaussianNoise(), ShotNoise(), ImpulseNoise(), SpeckleNoise()]
            self.blur = [GaussianBlur(), DefocusBlur(), MotionBlur(), GlassBlur(), ZoomBlur()]
            self.weather = [Fog(), Snow(), Frost(), Rain(), Shadow()]

            self.noises = [self.blur, self.noise, self.weather]
            self.processes = [self.camera, self.process]

            self.warp = [Curve(), Distort(), Stretch()]
            self.geometry = [Rotate(), Perspective(), Shrink()]

            self.isbaseline_aug = False
            # rand augment
            if self.opt.isrand_aug:
                self.augs = [self.process, self.camera, self.noise, self.blur, self.weather, self.pattern, self.warp, self.geometry]
            # semantic augment
            elif self.opt.issemantic_aug:
                self.geometry = [Rotate(), Perspective(), Shrink()]
                self.noise = [GaussianNoise()]
                self.blur = [MotionBlur()]
                self.augs = [self.noise, self.blur, self.geometry]
                self.isbaseline_aug = True
            # pp-ocr augment
            elif self.opt.islearning_aug:
                self.geometry = [Rotate(), Perspective()]
                self.noise = [GaussianNoise()]
                self.blur = [MotionBlur()]
                self.warp = [Distort()]
                self.augs = [self.warp, self.noise, self.blur, self.geometry]
                self.isbaseline_aug = True
            # scatter augment
            elif self.opt.isscatter_aug:
                self.geometry = [Shrink()]
                self.warp = [Distort()]
                self.augs = [self.warp, self.geometry]
                self.baseline_aug = True
            # rotation augment
            elif self.opt.isrotation_aug:
                self.geometry = [Rotate()]
                self.augs = [self.geometry]
                self.isbaseline_aug = True

        self.scale = False if opt.Transformer else True

    def __call__(self, img):
        '''
            Must call img.copy() if pattern, Rain or Shadow is used
        '''
        img = img.resize((self.opt.imgW, self.opt.imgH), Image.BICUBIC)

        if self.opt.eval or isless(self.opt.intact_prob):
            pass
        elif self.opt.isrand_aug or self.isbaseline_aug:
            img = self.rand_aug(img)
        # individual augment can also be selected
        elif self.opt.issel_aug:
            img = self.sel_aug(img)

        img = transforms.ToTensor()(img)
        if self.scale:
            img.sub_(0.5).div_(0.5)
        return img


    def rand_aug(self, img):
        augs = np.random.choice(self.augs, self.opt.augs_num, replace=False)
        for aug in augs:
            index = np.random.randint(0, len(aug))
            op = aug[index]
            mag = np.random.randint(0, 3) if self.opt.augs_mag is None else self.opt.augs_mag
            if type(op).__name__ == "Rain"  or type(op).__name__ == "Grid":
                img = op(img.copy(), mag=mag)
            else:
                img = op(img, mag=mag)

        return img

    def sel_aug(self, img):

        prob = 1.

        if self.opt.process:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.process))
            op = self.process[index]
            img = op(img, mag=mag, prob=prob)

        if self.opt.noise:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.noise))
            op = self.noise[index]
            img = op(img, mag=mag, prob=prob)

        if self.opt.blur:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.blur))
            op = self.blur[index]
            img = op(img, mag=mag, prob=prob)

        if self.opt.weather:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.weather))
            op = self.weather[index]
            if type(op).__name__ == "Rain": #or "Grid" in type(op).__name__ :
                img = op(img.copy(), mag=mag, prob=prob)
            else:
                img = op(img, mag=mag, prob=prob)

        if self.opt.camera:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.camera))
            op = self.camera[index]
            img = op(img, mag=mag, prob=prob)

        if self.opt.pattern:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.pattern))
            op = self.pattern[index]
            img = op(img.copy(), mag=mag, prob=prob)

        iscurve = False
        if self.opt.warp:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.warp))
            op = self.warp[index]
            if type(op).__name__ == "Curve":
                iscurve = True
            img = op(img, mag=mag, prob=prob)

        if self.opt.geometry:
            mag = np.random.randint(0, 3)
            index = np.random.randint(0, len(self.geometry))
            op = self.geometry[index]
            if type(op).__name__ == "Rotate":
                img = op(img, iscurve=iscurve, mag=mag, prob=prob)
            else:
                img = op(img, mag=mag, prob=prob)

        return img


class ResizeNormalize(object):

    def __init__(self, size, interpolation=Image.BICUBIC):
        self.size = size
        self.interpolation = interpolation
        self.toTensor = transforms.ToTensor()

    def __call__(self, img):
        img = img.resize(self.size, self.interpolation)
        img = self.toTensor(img)
        img.sub_(0.5).div_(0.5)
        return img


class NormalizePAD(object):

    def __init__(self, max_size, PAD_type='right'):
        self.toTensor = transforms.ToTensor()
        self.max_size = max_size
        self.max_width_half = math.floor(max_size[2] / 2)
        self.PAD_type = PAD_type

    def __call__(self, img):
        img = self.toTensor(img)
        img.sub_(0.5).div_(0.5)
        c, h, w = img.size()
        Pad_img = torch.FloatTensor(*self.max_size).fill_(0)
        Pad_img[:, :, :w] = img  # right pad
        if self.max_size[2] != w:  # add border Pad
            Pad_img[:, :, w:] = img[:, :, w - 1].unsqueeze(2).expand(c, h, self.max_size[2] - w)

        return Pad_img


class AlignCollate(object):

    def __init__(self, imgH=32, imgW=100, keep_ratio_with_pad=False, opt=None):
        self.imgH = imgH
        self.imgW = imgW
        self.keep_ratio_with_pad = keep_ratio_with_pad
        self.opt = opt

    def __call__(self, batch):
        batch = filter(lambda x: x is not None, batch)
        images, labels = zip(*batch)

        if self.keep_ratio_with_pad:  # same concept with 'Rosetta' paper
            resized_max_w = self.imgW
            input_channel = 3 if images[0].mode == 'RGB' else 1
            transform = NormalizePAD((input_channel, self.imgH, resized_max_w))

            resized_images = []
            for image in images:
                w, h = image.size
                ratio = w / float(h)
                if math.ceil(self.imgH * ratio) > self.imgW:
                    resized_w = self.imgW
                else:
                    resized_w = math.ceil(self.imgH * ratio)

                resized_image = image.resize((resized_w, self.imgH), Image.BICUBIC)
                resized_images.append(transform(resized_image))
                # resized_image.save('./image_test/%d_test.jpg' % w)

            image_tensors = torch.cat([t.unsqueeze(0) for t in resized_images], 0)

        else:
            transform = DataAugment(self.opt)
            #i = 0
            #for image in images:
            #    transform(image)
            #    if i == 1:
            #        exit(0)
            #    else: 
            #        i = i + 1
            image_tensors = [transform(image) for image in images]
            image_tensors = torch.cat([t.unsqueeze(0) for t in image_tensors], 0)
        
        #else:
        #    transform = ResizeNormalize((self.imgW, self.imgH))
        #    image_tensors = [transform(image) for image in images]
        #    image_tensors = torch.cat([t.unsqueeze(0) for t in image_tensors], 0)

        return image_tensors, labels


def tensor2im(image_tensor, imtype=np.uint8):
    image_numpy = image_tensor.cpu().float().numpy()
    if image_numpy.shape[0] == 1:
        image_numpy = np.tile(image_numpy, (3, 1, 1))
    image_numpy = (np.transpose(image_numpy, (1, 2, 0)) + 1) / 2.0 * 255.0
    return image_numpy.astype(imtype)


def save_image(image_numpy, image_path):
    image_pil = Image.fromarray(image_numpy)
    image_pil.save(image_path)