File size: 5,114 Bytes
6fc683c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
# from https://github.com/facebookresearch/detr/blob/main/d2/detr/dataset_mapper.py


import copy
import logging

import numpy as np
import torch

from detectron2.data import detection_utils as utils
from detectron2.data import transforms as T

from layoutlmft import LayoutLMv3Tokenizer

__all__ = ["DetrDatasetMapper"]


def build_transform_gen(cfg, is_train, aug_flip_crop=True):
    """
    Create a list of :class:`TransformGen` from config.
    Returns:
        list[TransformGen]
    """
    if is_train:
        min_size = cfg.INPUT.MIN_SIZE_TRAIN
        max_size = cfg.INPUT.MAX_SIZE_TRAIN
        sample_style = cfg.INPUT.MIN_SIZE_TRAIN_SAMPLING
    else:
        min_size = cfg.INPUT.MIN_SIZE_TEST
        max_size = cfg.INPUT.MAX_SIZE_TEST
        sample_style = "choice"
    if sample_style == "range":
        assert len(min_size) == 2, "more than 2 ({}) min_size(s) are provided for ranges".format(len(min_size))

    logger = logging.getLogger(__name__)
    tfm_gens = []
    if is_train and aug_flip_crop:
        tfm_gens.append(T.RandomFlip())
    tfm_gens.append(T.ResizeShortestEdge(min_size, max_size, sample_style))
    if is_train:
        logger.info("TransformGens used in training: " + str(tfm_gens))
    return tfm_gens


class DetrDatasetMapper:
    """
    A callable which takes a dataset dict in Detectron2 Dataset format,
    and map it into a format used by DETR.

    The callable currently does the following:

    1. Read the image from "file_name"
    2. Applies geometric transforms to the image and annotation
    3. Find and applies suitable cropping to the image and annotation
    4. Prepare image and annotation to Tensors
    """

    def __init__(self, cfg, is_train=True):
        self.img_format = cfg.INPUT.FORMAT
        self.is_train = is_train

        self.layoutlmv3 = 'layoutlmv3' in cfg.MODEL.VIT.NAME

        if self.layoutlmv3:
            # We disable the flipping/cropping augmentation in layoutlmv3 to be consistent with pre-training
            # Note that we do not disable resizing augmentation since the text boxes are also resized/normalized.
            aug_flip_crop = False
        else:
            aug_flip_crop = True

        if cfg.INPUT.CROP.ENABLED and is_train and aug_flip_crop:
            self.crop_gen = [
                T.ResizeShortestEdge([400, 500, 600], sample_style="choice"),
                T.RandomCrop(cfg.INPUT.CROP.TYPE, cfg.INPUT.CROP.SIZE),
            ]
        else:
            self.crop_gen = None

        self.mask_on = cfg.MODEL.MASK_ON
        self.tfm_gens = build_transform_gen(cfg, is_train, aug_flip_crop)
        logging.getLogger(__name__).info(
            "Full TransformGens used in training: {}, crop: {}".format(str(self.tfm_gens), str(self.crop_gen))
        )

    def __call__(self, dataset_dict):
        """
        Args:
            dataset_dict (dict): Metadata of one image, in Detectron2 Dataset format.

        Returns:
            dict: a format that builtin models in detectron2 accept
        """
        dataset_dict = copy.deepcopy(dataset_dict)  # it will be modified by code below

        image = utils.read_image(dataset_dict["file_name"], format=self.img_format)
        utils.check_image_size(dataset_dict, image)

        if self.crop_gen is None:
            image, transforms = T.apply_transform_gens(self.tfm_gens, image)
        else:
            if np.random.rand() > 0.5:
                image, transforms = T.apply_transform_gens(self.tfm_gens, image)
            else:
                image, transforms = T.apply_transform_gens(
                    self.tfm_gens[:-1] + self.crop_gen + self.tfm_gens[-1:], image
                )

        image_shape = image.shape[:2]  # h, w

        # Pytorch's dataloader is efficient on torch.Tensor due to shared-memory,
        # but not efficient on large generic data structures due to the use of pickle & mp.Queue.
        # Therefore it's important to use torch.Tensor.
        dataset_dict["image"] = torch.as_tensor(np.ascontiguousarray(image.transpose(2, 0, 1)))

        if not self.is_train:
            # USER: Modify this if you want to keep them for some reason.
            dataset_dict.pop("annotations", None)
            return dataset_dict

        if "annotations" in dataset_dict:
            # USER: Modify this if you want to keep them for some reason.
            for anno in dataset_dict["annotations"]:
                if not self.mask_on:
                    anno.pop("segmentation", None)
                anno.pop("keypoints", None)

            # USER: Implement additional transformations if you have other types of data
            annos = [
                utils.transform_instance_annotations(obj, transforms, image_shape)
                for obj in dataset_dict.pop("annotations")
                if obj.get("iscrowd", 0) == 0
            ]
            instances = utils.annotations_to_instances(annos, image_shape)
            dataset_dict["instances"] = utils.filter_empty_instances(instances)
        return dataset_dict