swin_transformer.py

# Copyright (c) 2021  PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License"
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from functools import lru_cache, reduce
from operator import mul

import numpy as np
import paddle
import paddle.nn as nn
import paddle.nn.functional as F
from paddle.nn.initializer import Constant

from ...utils import load_ckpt
from ..registry import BACKBONES
from ..weight_init import trunc_normal_

zeros_ = Constant(value=0.)
ones_ = Constant(value=1.)


def drop_path(x, drop_prob=0., training=False):
    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
    the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
    # issuecomment-532968956 ...
    See discussion: https://github.com/tensorflow/tpu/issues/494
    """
    if drop_prob == 0. or not training:
        return x
    keep_prob = paddle.to_tensor(1 - drop_prob)
    shape = (paddle.shape(x)[0], ) + (1, ) * (x.ndim - 1)
    random_tensor = keep_prob + paddle.rand(shape, dtype=x.dtype)
    random_tensor = paddle.floor(random_tensor)  # binarize
    output = x.divide(keep_prob) * random_tensor

    return output


class DropPath(nn.Layer):
    """Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks).
    """
    def __init__(self, drop_prob=None):
        super(DropPath, self).__init__()
        self.drop_prob = drop_prob

    def forward(self, x):
        return drop_path(x, self.drop_prob, self.training)


class Mlp(nn.Layer):
    """ Multilayer perceptron."""
    def __init__(self,
                 in_features,
                 hidden_features=None,
                 out_features=None,
                 act_layer=nn.GELU,
                 drop=0.):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)

    def forward(self, x):
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        x = self.drop(x)
        return x


def window_partition(x, window_size):
    """window_partition
    Args:
        x (Tensor): x.shape = [B, D, H, W, C]
        window_size (tuple[int]): window_size

    Returns:
        Tensor: (B*num_windows, window_size*window_size, C)
    """
    B, D, H, W, C = x.shape
    x = x.reshape([
        B, D // window_size[0], window_size[0], H // window_size[1],
        window_size[1], W // window_size[2], window_size[2], C
    ])
    windows = x.transpose([0, 1, 3, 5, 2, 4, 6,
                           7]).reshape([-1, reduce(mul, window_size), C])
    return windows


class Identity(nn.Layer):
    def __init__(self):
        super(Identity, self).__init__()

    def forward(self, input):
        return input


def window_reverse(windows, window_size, B, D, H, W):
    """
    Args:
        windows: (B*num_windows, window_size, window_size, C)
        window_size (tuple[int]): Window size
        H (int): Height of image
        W (int): Width of image

    Returns:
        x: (B, D, H, W, C)
    """
    x = windows.reshape([
        B, D // window_size[0], H // window_size[1], W // window_size[2],
        window_size[0], window_size[1], window_size[2], -1
    ])
    x = x.transpose([0, 1, 4, 2, 5, 3, 6, 7]).reshape([B, D, H, W, -1])
    return x


def get_window_size(x_size, window_size, shift_size=None):
    use_window_size = list(window_size)
    if shift_size is not None:
        use_shift_size = list(shift_size)
    for i in range(len(x_size)):
        if x_size[i] <= window_size[i]:
            use_window_size[i] = x_size[i]
            if shift_size is not None:
                use_shift_size[i] = 0

    if shift_size is None:
        return tuple(use_window_size)
    else:
        return tuple(use_window_size), tuple(use_shift_size)


class WindowAttention3D(nn.Layer):
    """ Window based multi-head self attention (W-MSA) module with relative position bias.
    It supports both of shifted and non-shifted window.
    Args:
        dim (int): Number of input channels.
        window_size (tuple[int]): The temporal length, height and width of the window.
        num_heads (int): Number of attention heads.
        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
    """
    def __init__(self,
                 dim,
                 window_size,
                 num_heads,
                 qkv_bias=False,
                 qk_scale=None,
                 attn_drop=0.,
                 proj_drop=0.):

        super().__init__()
        self.dim = dim
        self.window_size = window_size  # Wd, Wh, Ww
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim**-0.5

        # define a parameter table of relative position bias
        self.relative_position_bias_table = self.create_parameter(
            shape=((2 * window_size[0] - 1) * (2 * window_size[1] - 1) *
                   (2 * window_size[2] - 1), num_heads),
            default_initializer=zeros_,
        )  # 2*Wd-1 * 2*Wh-1 * 2*Ww-1, nH
        self.add_parameter("relative_position_bias_table",
                           self.relative_position_bias_table)
        # get pair-wise relative position index for each token inside the window
        coords_d = paddle.arange(self.window_size[0])
        coords_h = paddle.arange(self.window_size[1])
        coords_w = paddle.arange(self.window_size[2])
        coords = paddle.stack(paddle.meshgrid(coords_d, coords_h,
                                              coords_w))  # 3, Wd, Wh, Ww
        coords_flatten = paddle.flatten(coords, 1)  # 3, Wd*Wh*Ww

        relative_coords = coords_flatten.unsqueeze(
            axis=2) - coords_flatten.unsqueeze(axis=1)  # 3, Wd*Wh*Ww, Wd*Wh*Ww

        # relative_coords = coords_flatten.unsqueeze(2) - coords_flatten.unsqueeze(1)  # 3, Wd*Wh*Ww, Wd*Wh*Ww
        relative_coords = relative_coords.transpose([1, 2, 0
                                                     ])  # Wd*Wh*Ww, Wd*Wh*Ww, 3
        relative_coords[:, :,
                        0] += self.window_size[0] - 1  # shift to start from 0
        relative_coords[:, :, 1] += self.window_size[1] - 1
        relative_coords[:, :, 2] += self.window_size[2] - 1

        relative_coords[:, :, 0] *= (2 * self.window_size[1] -
                                     1) * (2 * self.window_size[2] - 1)
        relative_coords[:, :, 1] *= (2 * self.window_size[2] - 1)
        relative_position_index = relative_coords.sum(
            axis=-1)  # Wd*Wh*Ww, Wd*Wh*Ww
        self.register_buffer("relative_position_index", relative_position_index)

        self.qkv = nn.Linear(dim, dim * 3, bias_attr=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

        trunc_normal_(self.relative_position_bias_table, std=0.02)
        self.softmax = nn.Softmax(axis=-1)

    def forward(self, x, mask=None):
        """ Forward function.
        Args:
            x: input features with shape of (num_windows*B, N, C)
            mask: (0/-inf) mask with shape of (num_windows, N, N) or None
        """
        B_, N, C = x.shape
        qkv = self.qkv(x).reshape(
            [B_, N, 3, self.num_heads,
             C // self.num_heads]).transpose([2, 0, 3, 1, 4])
        q, k, v = qkv[0], qkv[1], qkv[2]  # B_, nH, N, C

        q = q * self.scale
        attn = q @ k.transpose([0, 1, 3, 2])

        relative_position_bias = self.relative_position_bias_table[
            self.relative_position_index[:N, :N].reshape([-1])].reshape(
                [N, N, -1])  # Wd*Wh*Ww,Wd*Wh*Ww,nH
        relative_position_bias = relative_position_bias.transpose(
            [2, 0, 1])  # nH, Wd*Wh*Ww, Wd*Wh*Ww
        attn = attn + relative_position_bias.unsqueeze(0)  # B_, nH, N, N

        if mask is not None:
            nW = mask.shape[0]
            attn = attn.reshape([B_ // nW, nW, self.num_heads, N, N
                                 ]) + mask.unsqueeze(1).unsqueeze(0)
            attn = attn.reshape([-1, self.num_heads, N, N])
            attn = self.softmax(attn)
        else:
            attn = self.softmax(attn)

        attn = self.attn_drop(attn)

        x = (attn @ v).transpose([0, 2, 1, 3]).reshape([B_, N, C])
        x = self.proj(x)
        x = self.proj_drop(x)
        return x


class SwinTransformerBlock3D(nn.Layer):
    """ Swin Transformer Block.

    Args:
        dim (int): Number of input channels.
        num_heads (int): Number of attention heads.
        window_size (tuple[int]): Window size.
        shift_size (tuple[int]): Shift size for SW-MSA.
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
        drop (float, optional): Dropout rate. Default: 0.0
        attn_drop (float, optional): Attention dropout rate. Default: 0.0
        drop_path (float, optional): Stochastic depth rate. Default: 0.0
        act_layer (nn.Layer, optional): Activation layer. Default: nn.GELU
        norm_layer (nn.Layer, optional): Normalization layer.  Default: nn.LayerNorm
    """
    def __init__(self,
                 dim,
                 num_heads,
                 window_size=(2, 7, 7),
                 shift_size=(0, 0, 0),
                 mlp_ratio=4.,
                 qkv_bias=True,
                 qk_scale=None,
                 drop=0.,
                 attn_drop=0.,
                 drop_path=0.,
                 act_layer=nn.GELU,
                 norm_layer=nn.LayerNorm,
                 use_checkpoint=False):
        super().__init__()
        self.dim = dim
        self.num_heads = num_heads
        self.window_size = window_size
        self.shift_size = shift_size
        self.mlp_ratio = mlp_ratio
        # self.use_checkpoint=use_checkpoint

        assert 0 <= self.shift_size[0] < self.window_size[
            0], "shift_size must in 0-window_size"
        assert 0 <= self.shift_size[1] < self.window_size[
            1], "shift_size must in 0-window_size"
        assert 0 <= self.shift_size[2] < self.window_size[
            2], "shift_size must in 0-window_size"

        self.norm1 = norm_layer(dim)
        self.attn = WindowAttention3D(dim,
                                      window_size=self.window_size,
                                      num_heads=num_heads,
                                      qkv_bias=qkv_bias,
                                      qk_scale=qk_scale,
                                      attn_drop=attn_drop,
                                      proj_drop=drop)

        self.drop_path = DropPath(drop_path) if drop_path > 0. else Identity()
        self.norm2 = norm_layer(dim)
        mlp_hidden_dim = int(dim * mlp_ratio)
        self.mlp = Mlp(in_features=dim,
                       hidden_features=mlp_hidden_dim,
                       act_layer=act_layer,
                       drop=drop)

    def forward_part1(self, x, mask_matrix):
        B = paddle.shape(x)[0]
        _, D, H, W, C = x.shape
        window_size, shift_size = get_window_size((D, H, W), self.window_size,
                                                  self.shift_size)

        x = self.norm1(x)
        # pad feature maps to multiples of window size
        pad_l = pad_t = pad_d0 = 0
        pad_d1 = (window_size[0] - D % window_size[0]) % window_size[0]
        pad_b = (window_size[1] - H % window_size[1]) % window_size[1]
        pad_r = (window_size[2] - W % window_size[2]) % window_size[2]
        x = F.pad(x, (pad_l, pad_r, pad_t, pad_b, pad_d0, pad_d1),
                  data_format='NDHWC')
        _, Dp, Hp, Wp, _ = x.shape
        # cyclic shift
        if any(i > 0 for i in shift_size):
            shifted_x = paddle.roll(x,
                                    shifts=(-shift_size[0], -shift_size[1],
                                            -shift_size[2]),
                                    axis=(1, 2, 3))
            attn_mask = mask_matrix
        else:
            shifted_x = x
            attn_mask = None
        # partition windows
        x_windows = window_partition(shifted_x,
                                     window_size)  # B*nW, Wd*Wh*Ww, C
        # W-MSA/SW-MSA
        attn_windows = self.attn(x_windows, mask=attn_mask)  # B*nW, Wd*Wh*Ww, C
        # merge windows
        attn_windows = attn_windows.reshape([-1, *(window_size + (C, ))])
        shifted_x = window_reverse(attn_windows, window_size, B, Dp, Hp,
                                   Wp)  # B D' H' W' C
        # reverse cyclic shift
        if any(i > 0 for i in shift_size):
            x = paddle.roll(shifted_x,
                            shifts=(shift_size[0], shift_size[1],
                                    shift_size[2]),
                            axis=(1, 2, 3))
        else:
            x = shifted_x

        if pad_d1 > 0 or pad_r > 0 or pad_b > 0:
            x = x[:, :D, :H, :W, :]
        return x

    def forward_part2(self, x):
        return self.drop_path(self.mlp(self.norm2(x)))

    def forward(self, x, mask_matrix):
        """ Forward function.

        Args:
            x: Input feature, tensor size (B, D, H, W, C).
            mask_matrix: Attention mask for cyclic shift.
        """

        shortcut = x
        x = self.forward_part1(x, mask_matrix)
        x = shortcut + self.drop_path(x)
        x = x + self.forward_part2(x)

        return x


class PatchMerging(nn.Layer):
    """ Patch Merging Layer

    Args:
        dim (int): Number of input channels.
        norm_layer (nn.Layer, optional): Normalization layer.  Default: nn.LayerNorm
    """
    def __init__(self, dim, norm_layer=nn.LayerNorm):
        super().__init__()
        self.dim = dim
        self.reduction = nn.Linear(4 * dim, 2 * dim, bias_attr=False)
        self.norm = norm_layer(4 * dim)

    def forward(self, x):
        """ Forward function.

        Args:
            x: Input feature, tensor size (B, D, H, W, C).
        """
        B, D, H, W, C = x.shape

        # padding
        pad_input = (H % 2 == 1) or (W % 2 == 1)
        if pad_input:
            x = F.pad(x, (0, W % 2, 0, H % 2, 0, 0), data_format='NDHWC')

        x0 = x[:, :, 0::2, 0::2, :]  # B D H/2 W/2 C
        x1 = x[:, :, 1::2, 0::2, :]  # B D H/2 W/2 C
        x2 = x[:, :, 0::2, 1::2, :]  # B D H/2 W/2 C
        x3 = x[:, :, 1::2, 1::2, :]  # B D H/2 W/2 C
        x = paddle.concat([x0, x1, x2, x3], -1)  # B D H/2 W/2 4*C

        x = self.norm(x)
        x = self.reduction(x)

        return x


# cache each stage results
@lru_cache()
def compute_mask(D, H, W, window_size, shift_size):
    img_mask = paddle.zeros((1, D, H, W, 1))  # 1 Dp Hp Wp 1
    cnt = 0
    for d in slice(-window_size[0]), slice(-window_size[0],
                                           -shift_size[0]), slice(
                                               -shift_size[0], None):
        for h in slice(-window_size[1]), slice(-window_size[1],
                                               -shift_size[1]), slice(
                                                   -shift_size[1], None):
            for w in slice(-window_size[2]), slice(-window_size[2],
                                                   -shift_size[2]), slice(
                                                       -shift_size[2], None):
                img_mask[:, d, h, w, :] = cnt
                cnt += 1
    mask_windows = window_partition(img_mask,
                                    window_size)  # nW, ws[0]*ws[1]*ws[2], 1
    mask_windows = mask_windows.squeeze(-1)  # nW, ws[0]*ws[1]*ws[2]
    attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
    # attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
    huns = -100.0 * paddle.ones_like(attn_mask)
    attn_mask = huns * (attn_mask != 0).astype("float32")
    return attn_mask


class BasicLayer(nn.Layer):
    """ A basic Swin Transformer layer for one stage.

    Args:
        dim (int): Number of feature channels
        depth (int): Depths of this stage.
        num_heads (int): Number of attention head.
        window_size (tuple[int]): Local window size. Default: (1,7,7).
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
        drop (float, optional): Dropout rate. Default: 0.0
        attn_drop (float, optional): Attention dropout rate. Default: 0.0
        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
        norm_layer (nn.Layer, optional): Normalization layer. Default: nn.LayerNorm
        downsample (nn.Layer | None, optional): Downsample layer at the end of the layer. Default: None
    """
    def __init__(self,
                 dim,
                 depth,
                 num_heads,
                 window_size=(1, 7, 7),
                 mlp_ratio=4.,
                 qkv_bias=False,
                 qk_scale=None,
                 drop=0.,
                 attn_drop=0.,
                 drop_path=0.,
                 norm_layer=nn.LayerNorm,
                 downsample=None,
                 use_checkpoint=False):
        super().__init__()
        self.window_size = window_size
        self.shift_size = tuple(i // 2 for i in window_size)
        self.depth = depth
        self.use_checkpoint = use_checkpoint

        # build blocks
        self.blocks = nn.LayerList([
            SwinTransformerBlock3D(
                dim=dim,
                num_heads=num_heads,
                window_size=window_size,
                shift_size=(0, 0, 0) if (i % 2 == 0) else self.shift_size,
                mlp_ratio=mlp_ratio,
                qkv_bias=qkv_bias,
                qk_scale=qk_scale,
                drop=drop,
                attn_drop=attn_drop,
                drop_path=drop_path[i]
                if isinstance(drop_path, list) else drop_path,
                norm_layer=norm_layer,
                use_checkpoint=use_checkpoint,
            ) for i in range(depth)
        ])

        self.downsample = downsample
        if self.downsample is not None:
            self.downsample = downsample(dim=dim, norm_layer=norm_layer)

    def forward(self, x):
        """ Forward function.

        Args:
            x: Input feature, tensor size (B, C, D, H, W).
        """
        # calculate attention mask for SW-MSA
        B = paddle.shape(x)[0]
        _, C, D, H, W = x.shape
        window_size, shift_size = get_window_size((D, H, W), self.window_size,
                                                  self.shift_size)
        # x = rearrange(x, 'b c d h w -> b d h w c')
        x = x.transpose([0, 2, 3, 4, 1])
        Dp = int(np.ceil(D / window_size[0])) * window_size[0]
        Hp = int(np.ceil(H / window_size[1])) * window_size[1]
        Wp = int(np.ceil(W / window_size[2])) * window_size[2]
        attn_mask = compute_mask(Dp, Hp, Wp, window_size, shift_size)
        for blk in self.blocks:
            x = blk(x, attn_mask)
        x = x.reshape([B, D, H, W, C])

        if self.downsample is not None:
            x = self.downsample(x)
        x = x.transpose([0, 4, 1, 2, 3])
        return x


class PatchEmbed3D(nn.Layer):
    """ Video to Patch Embedding.

    Args:
        patch_size (int): Patch token size. Default: (2,4,4).
        in_chans (int): Number of input video channels. Default: 3.
        embed_dim (int): Number of linear projection output channels. Default: 96.
        norm_layer (nn.Layer, optional): Normalization layer. Default: None
    """
    def __init__(self,
                 patch_size=(2, 4, 4),
                 in_chans=3,
                 embed_dim=96,
                 norm_layer=None):
        super().__init__()
        self.patch_size = patch_size

        self.in_chans = in_chans
        self.embed_dim = embed_dim

        self.proj = nn.Conv3D(in_chans,
                              embed_dim,
                              kernel_size=patch_size,
                              stride=patch_size)
        if norm_layer is not None:
            self.norm = norm_layer(embed_dim)
        else:
            self.norm = None

    def forward(self, x):
        _, _, D, H, W = x.shape
        if W % self.patch_size[2] != 0:
            x = F.pad(
                x, (0, self.patch_size[2] - W % self.patch_size[2], 0, 0, 0, 0),
                data_format='NCDHW')
        if H % self.patch_size[1] != 0:
            x = F.pad(
                x, (0, 0, 0, self.patch_size[1] - H % self.patch_size[1], 0, 0),
                data_format='NCDHW')
        if D % self.patch_size[0] != 0:
            x = F.pad(
                x, (0, 0, 0, 0, 0, self.patch_size[0] - D % self.patch_size[0]),
                data_format='NCDHW')

        x = self.proj(x)  # B C D Wh Ww
        if self.norm is not None:
            D, Wh, Ww = x.shape[2], x.shape[3], x.shape[4]
            x = x.flatten(2).transpose([0, 2, 1])
            x = self.norm(x)
            x = x.transpose([0, 2, 1]).reshape([-1, self.embed_dim, D, Wh, Ww])

        return x


@BACKBONES.register()
class SwinTransformer3D(nn.Layer):
    """ Swin Transformer backbone.
        A Paddle impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
          https://arxiv.org/pdf/2103.14030

    Args:
        patch_size (int | tuple(int)): Patch size. Default: (4,4,4).
        in_chans (int): Number of input image channels. Default: 3.
        embed_dim (int): Number of linear projection output channels. Default: 96.
        depths (tuple[int]): Depths of each Swin Transformer stage.
        num_heads (tuple[int]): Number of attention head of each stage.
        window_size (int): Window size. Default: 7.
        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: Truee
        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set.
        drop_rate (float): Dropout rate.
        attn_drop_rate (float): Attention dropout rate. Default: 0.
        drop_path_rate (float): Stochastic depth rate. Default: 0.2.
        norm_layer: Normalization layer. Default: nn.LayerNorm.
        patch_norm (bool): If True, add normalization after patch embedding. Default: False.
        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
            -1 means not freezing any parameters.
    """
    def __init__(self,
                 pretrained=None,
                 patch_size=(4, 4, 4),
                 in_chans=3,
                 embed_dim=96,
                 depths=[2, 2, 6, 2],
                 num_heads=[3, 6, 12, 24],
                 window_size=(2, 7, 7),
                 mlp_ratio=4.,
                 qkv_bias=True,
                 qk_scale=None,
                 drop_rate=0.,
                 attn_drop_rate=0.,
                 drop_path_rate=0.2,
                 norm_layer=nn.LayerNorm,
                 patch_norm=False,
                 frozen_stages=-1,
                 use_checkpoint=False):
        super().__init__()

        self.pretrained = pretrained
        self.num_layers = len(depths)
        self.embed_dim = embed_dim
        self.patch_norm = patch_norm
        self.frozen_stages = frozen_stages
        self.window_size = window_size
        self.patch_size = patch_size

        # split image into non-overlapping patches
        self.patch_embed = PatchEmbed3D(
            patch_size=patch_size,
            in_chans=in_chans,
            embed_dim=embed_dim,
            norm_layer=norm_layer if self.patch_norm else None)

        self.pos_drop = nn.Dropout(p=drop_rate)

        # stochastic depth
        dpr = [
            x.item() for x in paddle.linspace(0, drop_path_rate, sum(depths))
        ]  # stochastic depth decay rule

        # build layers
        self.layers = nn.LayerList()
        for i_layer in range(self.num_layers):
            layer = BasicLayer(
                dim=int(embed_dim * 2**i_layer),
                depth=depths[i_layer],
                num_heads=num_heads[i_layer],
                window_size=window_size,
                mlp_ratio=mlp_ratio,
                qkv_bias=qkv_bias,
                qk_scale=qk_scale,
                drop=drop_rate,
                attn_drop=attn_drop_rate,
                drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],
                norm_layer=norm_layer,
                downsample=PatchMerging
                if i_layer < self.num_layers - 1 else None,
                use_checkpoint=use_checkpoint)
            self.layers.append(layer)

        self.num_features = int(embed_dim * 2**(self.num_layers - 1))

        # add a norm layer for each output
        self.norm = norm_layer(self.num_features)

        self._freeze_stages()

    def _freeze_stages(self):
        if self.frozen_stages >= 0:
            self.patch_embed.eval()
            for param in self.patch_embed.parameters():
                param.stop_gradient = True

        if self.frozen_stages >= 1:
            self.pos_drop.eval()
            for i in range(0, self.frozen_stages):
                m = self.layers[i]
                m.eval()
                for param in m.parameters():
                    param.stop_gradient = True

    def _init_fn(self, m):
        if isinstance(m, nn.Linear):
            trunc_normal_(m.weight, std=0.02)
            if m.bias is not None:
                zeros_(m.bias)
        elif isinstance(m, nn.LayerNorm):
            zeros_(m.bias)
            ones_(m.weight)

    def init_weights(self):
        """Initialize the weights in backbone.

        Args:
            pretrained (str, optional): Path to pre-trained weights.
                Defaults to None.
        """
        """First init model's weight"""

        self.apply(self._init_fn)
        """Second, if provide pretrained ckpt, load it"""
        if isinstance(
                self.pretrained, str
        ) and self.pretrained.strip() != "":  # load pretrained weights
            load_ckpt(self, self.pretrained)
        elif self.pretrained is None or self.pretrained.strip() == "":
            pass
        else:
            raise NotImplementedError

    def forward(self, x):
        """Forward function."""
        x = self.patch_embed(x)
        x = self.pos_drop(x)

        for layer in self.layers:
            x = layer(x)

        x = x.transpose([0, 2, 3, 4, 1])
        x = self.norm(x)
        x = x.transpose([0, 4, 1, 2, 3])
        return x

    def train(self, mode=True):
        """Convert the model into training mode while keep layers freezed."""
        super(SwinTransformer3D, self).train(mode)
        self._freeze_stages()