From e3741de333a3a43a7968241b6eccaaac66dd7b20 Mon Sep 17 00:00:00 2001 From: Gustaf Rydholm Date: Sun, 21 Mar 2021 22:33:58 +0100 Subject: Working on EMNIST Lines dataset --- text_recognizer/tests/support/__init__.py | 2 - .../support/create_emnist_lines_support_files.py | 51 --------------------- .../tests/support/create_emnist_support_files.py | 30 ------------ .../support/create_iam_lines_support_files.py | 50 -------------------- .../tests/support/emnist_lines/Knox Ky.png | Bin 2301 -> 0 bytes .../emnist_lines/ancillary beliefs and.png | Bin 5424 -> 0 bytes .../tests/support/emnist_lines/they.png | Bin 1391 -> 0 bytes .../He rose from his breakfast-nook bench.png | Bin 5170 -> 0 bytes .../and came into the livingroom, where.png | Bin 3617 -> 0 bytes .../his entrance. He came, almost falling.png | Bin 3923 -> 0 bytes .../tests/support/iam_paragraphs/a01-000u.jpg | Bin 14890 -> 0 bytes 11 files changed, 133 deletions(-) delete mode 100644 text_recognizer/tests/support/__init__.py delete mode 100644 text_recognizer/tests/support/create_emnist_lines_support_files.py delete mode 100644 text_recognizer/tests/support/create_emnist_support_files.py delete mode 100644 text_recognizer/tests/support/create_iam_lines_support_files.py delete mode 100644 text_recognizer/tests/support/emnist_lines/Knox Ky.png delete mode 100644 text_recognizer/tests/support/emnist_lines/ancillary beliefs and.png delete mode 100644 text_recognizer/tests/support/emnist_lines/they.png delete mode 100644 text_recognizer/tests/support/iam_lines/He rose from his breakfast-nook bench.png delete mode 100644 text_recognizer/tests/support/iam_lines/and came into the livingroom, where.png delete mode 100644 text_recognizer/tests/support/iam_lines/his entrance. He came, almost falling.png delete mode 100644 text_recognizer/tests/support/iam_paragraphs/a01-000u.jpg (limited to 'text_recognizer/tests/support') diff --git a/text_recognizer/tests/support/__init__.py b/text_recognizer/tests/support/__init__.py deleted file mode 100644 index a265ede..0000000 --- a/text_recognizer/tests/support/__init__.py +++ /dev/null @@ -1,2 +0,0 @@ -"""Support file modules.""" -from .create_emnist_support_files import create_emnist_support_files diff --git a/text_recognizer/tests/support/create_emnist_lines_support_files.py b/text_recognizer/tests/support/create_emnist_lines_support_files.py deleted file mode 100644 index 9abe143..0000000 --- a/text_recognizer/tests/support/create_emnist_lines_support_files.py +++ /dev/null @@ -1,51 +0,0 @@ -"""Module for creating EMNIST Lines test support files.""" -# flake8: noqa: S106 - -from pathlib import Path -import shutil - -import numpy as np - -from text_recognizer.datasets import EmnistLinesDataset -import text_recognizer.util as util - - -SUPPORT_DIRNAME = Path(__file__).parents[0].resolve() / "emnist_lines" - - -def create_emnist_lines_support_files() -> None: - """Create EMNIST Lines test images.""" - shutil.rmtree(SUPPORT_DIRNAME, ignore_errors=True) - SUPPORT_DIRNAME.mkdir() - - # TODO: maybe have to add args to dataset. - dataset = EmnistLinesDataset( - init_token="", - pad_token="_", - eos_token="", - transform=[{"type": "ToTensor", "args": {}}], - target_transform=[ - { - "type": "AddTokens", - "args": {"init_token": "", "pad_token": "_", "eos_token": ""}, - } - ], - ) # nosec: S106 - dataset.load_or_generate_data() - - for index in [5, 7, 9]: - image, target = dataset[index] - if len(image.shape) == 3: - image = image.squeeze(0) - print(image.sum(), image.dtype) - - label = "".join(dataset.mapper(label) for label in target[1:]).strip( - dataset.mapper.pad_token - ) - print(label) - image = image.numpy() - util.write_image(image, str(SUPPORT_DIRNAME / f"{label}.png")) - - -if __name__ == "__main__": - create_emnist_lines_support_files() diff --git a/text_recognizer/tests/support/create_emnist_support_files.py b/text_recognizer/tests/support/create_emnist_support_files.py deleted file mode 100644 index f9ff030..0000000 --- a/text_recognizer/tests/support/create_emnist_support_files.py +++ /dev/null @@ -1,30 +0,0 @@ -"""Module for creating EMNIST test support files.""" -from pathlib import Path -import shutil - -from text_recognizer.datasets import EmnistDataset -from text_recognizer.util import write_image - -SUPPORT_DIRNAME = Path(__file__).parents[0].resolve() / "emnist" - - -def create_emnist_support_files() -> None: - """Create support images for test of CharacterPredictor class.""" - shutil.rmtree(SUPPORT_DIRNAME, ignore_errors=True) - SUPPORT_DIRNAME.mkdir() - - dataset = EmnistDataset(train=False) - dataset.load_or_generate_data() - - for index in [5, 7, 9]: - image, label = dataset[index] - if len(image.shape) == 3: - image = image.squeeze(0) - image = image.numpy() - label = dataset.mapper(int(label)) - print(index, label) - write_image(image, str(SUPPORT_DIRNAME / f"{label}.png")) - - -if __name__ == "__main__": - create_emnist_support_files() diff --git a/text_recognizer/tests/support/create_iam_lines_support_files.py b/text_recognizer/tests/support/create_iam_lines_support_files.py deleted file mode 100644 index 50f9e3d..0000000 --- a/text_recognizer/tests/support/create_iam_lines_support_files.py +++ /dev/null @@ -1,50 +0,0 @@ -"""Module for creating IAM Lines test support files.""" -# flake8: noqa -from pathlib import Path -import shutil - -import numpy as np - -from text_recognizer.datasets import IamLinesDataset -import text_recognizer.util as util - - -SUPPORT_DIRNAME = Path(__file__).parents[0].resolve() / "iam_lines" - - -def create_emnist_lines_support_files() -> None: - """Create IAM Lines test images.""" - shutil.rmtree(SUPPORT_DIRNAME, ignore_errors=True) - SUPPORT_DIRNAME.mkdir() - - # TODO: maybe have to add args to dataset. - dataset = IamLinesDataset( - init_token="", - pad_token="_", - eos_token="", - transform=[{"type": "ToTensor", "args": {}}], - target_transform=[ - { - "type": "AddTokens", - "args": {"init_token": "", "pad_token": "_", "eos_token": ""}, - } - ], - ) - dataset.load_or_generate_data() - - for index in [0, 1, 3]: - image, target = dataset[index] - if len(image.shape) == 3: - image = image.squeeze(0) - print(image.sum(), image.dtype) - - label = "".join(dataset.mapper(label) for label in target[1:]).strip( - dataset.mapper.pad_token - ) - print(label) - image = image.numpy() - util.write_image(image, str(SUPPORT_DIRNAME / f"{label}.png")) - - -if __name__ == "__main__": - create_emnist_lines_support_files() diff --git a/text_recognizer/tests/support/emnist_lines/Knox Ky.png b/text_recognizer/tests/support/emnist_lines/Knox Ky.png deleted file mode 100644 index b7d0618..0000000 Binary files a/text_recognizer/tests/support/emnist_lines/Knox Ky.png and /dev/null differ diff --git a/text_recognizer/tests/support/emnist_lines/ancillary beliefs and.png b/text_recognizer/tests/support/emnist_lines/ancillary beliefs and.png deleted file mode 100644 index 14a8cf3..0000000 Binary files a/text_recognizer/tests/support/emnist_lines/ancillary beliefs and.png and /dev/null differ diff --git a/text_recognizer/tests/support/emnist_lines/they.png b/text_recognizer/tests/support/emnist_lines/they.png deleted file mode 100644 index 7f05951..0000000 Binary files a/text_recognizer/tests/support/emnist_lines/they.png and /dev/null differ diff --git a/text_recognizer/tests/support/iam_lines/He rose from his breakfast-nook bench.png b/text_recognizer/tests/support/iam_lines/He rose from his breakfast-nook bench.png deleted file mode 100644 index 6eeb642..0000000 Binary files a/text_recognizer/tests/support/iam_lines/He rose from his breakfast-nook bench.png and /dev/null differ diff --git a/text_recognizer/tests/support/iam_lines/and came into the livingroom, where.png b/text_recognizer/tests/support/iam_lines/and came into the livingroom, where.png deleted file mode 100644 index 4974cf8..0000000 Binary files a/text_recognizer/tests/support/iam_lines/and came into the livingroom, where.png and /dev/null differ diff --git a/text_recognizer/tests/support/iam_lines/his entrance. He came, almost falling.png b/text_recognizer/tests/support/iam_lines/his entrance. He came, almost falling.png deleted file mode 100644 index a731245..0000000 Binary files a/text_recognizer/tests/support/iam_lines/his entrance. He came, almost falling.png and /dev/null differ diff --git a/text_recognizer/tests/support/iam_paragraphs/a01-000u.jpg b/text_recognizer/tests/support/iam_paragraphs/a01-000u.jpg deleted file mode 100644 index d9753b6..0000000 Binary files a/text_recognizer/tests/support/iam_paragraphs/a01-000u.jpg and /dev/null differ -- cgit v1.2.3-70-g09d2