pytorch-image-models/timm/models/layers/conv2d_same.py

""" Conv2d w/ Same Padding

Hacked together by Ross Wightman
"""
import torch
import torch.nn as nn
import torch.nn.functional as F
from typing import Union, List, Tuple, Optional, Callable
import math

from .padding import get_padding, pad_same, is_static_pad


def conv2d_same(
        x, weight: torch.Tensor, bias: Optional[torch.Tensor] = None, stride: Tuple[int, int] = (1, 1),
        padding: Tuple[int, int] = (0, 0), dilation: Tuple[int, int] = (1, 1), groups: int = 1):
    x = pad_same(x, weight.shape[-2:], stride, dilation)
    return F.conv2d(x, weight, bias, stride, (0, 0), dilation, groups)


class Conv2dSame(nn.Conv2d):
    """ Tensorflow like 'SAME' convolution wrapper for 2D convolutions
    """

    def __init__(self, in_channels, out_channels, kernel_size, stride=1,
                 padding=0, dilation=1, groups=1, bias=True):
        super(Conv2dSame, self).__init__(
            in_channels, out_channels, kernel_size, stride, 0, dilation, groups, bias)

    def forward(self, x):
        return conv2d_same(x, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups)


def get_padding_value(padding, kernel_size, **kwargs) -> Tuple[Tuple, bool]:
    dynamic = False
    if isinstance(padding, str):
        # for any string padding, the padding will be calculated for you, one of three ways
        padding = padding.lower()
        if padding == 'same':
            # TF compatible 'SAME' padding, has a performance and GPU memory allocation impact
            if is_static_pad(kernel_size, **kwargs):
                # static case, no extra overhead
                padding = get_padding(kernel_size, **kwargs)
            else:
                # dynamic 'SAME' padding, has runtime/GPU memory overhead
                padding = 0
                dynamic = True
        elif padding == 'valid':
            # 'VALID' padding, same as padding=0
            padding = 0
        else:
            # Default to PyTorch style 'same'-ish symmetric padding
            padding = get_padding(kernel_size, **kwargs)
    return padding, dynamic


def create_conv2d_pad(in_chs, out_chs, kernel_size, **kwargs):
    padding = kwargs.pop('padding', '')
    kwargs.setdefault('bias', False)
    padding, is_dynamic = get_padding_value(padding, kernel_size, **kwargs)
    if is_dynamic:
        return Conv2dSame(in_chs, out_chs, kernel_size, **kwargs)
    else:
        return nn.Conv2d(in_chs, out_chs, kernel_size, padding=padding, **kwargs)