johschmidt42

# Imports
import pathlib

import numpy as np
import torch
from skimage.io import imread
from skimage.transform import resize

from inference import predict
from transformations import normalize_01, re_normalize

johschmidt42

import torch
from skimage.io import imread
from torch.utils import data


class SegmentationDataSet(data.Dataset):
    def __init__(self,
                 inputs: list,
                 targets: list,
                 transform=None,

johschmidt42

from torch import nn
import torch


@torch.jit.script
def autocrop(encoder_layer: torch.Tensor, decoder_layer: torch.Tensor):
    """
    Center-crops the encoder_layer to the size of the decoder_layer,
    so that merging (concatenation) between levels/blocks is possible.
    This is only necessary for input sizes != 2**n for 'same' padding and always required for 'valid' padding.

johschmidt42

import pathlib
from utils import get_filenames_of_path

root = pathlib.Path(".../Heads")

inputs = get_filenames_of_path(root / 'input')
inputs.sort()


for idx, path in enumerate(inputs):

johschmidt42

import pathlib
from typing import List

import numpy as np

from pytorch_faster_rcnn_tutorial.annotator import Annotator
from pytorch_faster_rcnn_tutorial.datasets import ObjectDetectionDatasetSingle
from pytorch_faster_rcnn_tutorial.transformations import (
    ComposeSingle,
    FunctionWrapperSingle,

johschmidt42

import pathlib
import torch
from utils import get_filenames_of_path

root = pathlib.Path("...\Heads")

targets = get_filenames_of_path(root / 'target')
targets.sort()

annotation = torch.load(targets[1])

johschmidt42

import logging
import pathlib
import sys
import warnings
from typing import List

import numpy as np

from pytorch_faster_rcnn_tutorial.datasets import ObjectDetectionDataSet
from pytorch_faster_rcnn_tutorial.transformations import (

johschmidt42

import logging
import pathlib
from multiprocessing import Pool
from typing import Dict, List

import torch
from skimage.color import rgba2rgb
from skimage.io import imread
from torch.utils.data import Dataset
from torchvision.ops import box_convert

johschmidt42

# visualize dataset

color_mapping = {
    1: 'red',
}

from visual import DatasetViewer

datasetviewer = DatasetViewer(dataset, color_mapping)
datasetviewer.napari()

johschmidt42

from pytorch_faster_rcnn_tutorial.viewers.object_detection_viewer import ObjectDetectionViewer
from torchvision.models.detection.transform import GeneralizedRCNNTransform

color_mapping: Dict[int, str] = {
    1: "red",
}

transform: GeneralizedRCNNTransform = GeneralizedRCNNTransform(
    min_size=1024,
    max_size=1024,