modeling_basnet.py

import logging
from dataclasses import dataclass
from typing import Optional, Tuple, Union

import torch
import torch.nn as nn
import torchvision
from transformers.modeling_utils import PreTrainedModel
from transformers.utils import ModelOutput

from .configuration_basnet import BASNetConfig

logger = logging.getLogger(__name__)


@dataclass
class BasNetSideOutput(ModelOutput):
    dout: torch.Tensor
    d1: Optional[torch.Tensor] = None
    d2: Optional[torch.Tensor] = None
    d3: Optional[torch.Tensor] = None
    d4: Optional[torch.Tensor] = None
    d5: Optional[torch.Tensor] = None
    d6: Optional[torch.Tensor] = None
    db: Optional[torch.Tensor] = None


@dataclass
class BASNetModelOutput(ModelOutput):
    activated: BasNetSideOutput


class RefUnet(nn.Module):
    def __init__(self, in_ch: int, inc_ch: int) -> None:
        super().__init__()

        self.conv0 = nn.Conv2d(in_ch, inc_ch, kernel_size=3, padding=1)

        self.conv1 = nn.Conv2d(inc_ch, 64, kernel_size=3, padding=1)
        self.bn1 = nn.BatchNorm2d(64)
        self.relu1 = nn.ReLU(inplace=True)

        self.pool1 = nn.MaxPool2d(2, 2, ceil_mode=True)

        self.conv2 = nn.Conv2d(64, 64, kernel_size=3, padding=1)
        self.bn2 = nn.BatchNorm2d(64)
        self.relu2 = nn.ReLU(inplace=True)

        self.pool2 = nn.MaxPool2d(2, 2, ceil_mode=True)

        self.conv3 = nn.Conv2d(64, 64, kernel_size=3, padding=1)
        self.bn3 = nn.BatchNorm2d(64)
        self.relu3 = nn.ReLU(inplace=True)

        self.pool3 = nn.MaxPool2d(2, 2, ceil_mode=True)

        self.conv4 = nn.Conv2d(64, 64, kernel_size=3, padding=1)
        self.bn4 = nn.BatchNorm2d(64)
        self.relu4 = nn.ReLU(inplace=True)

        self.pool4 = nn.MaxPool2d(2, 2, ceil_mode=True)

        #####

        self.conv5 = nn.Conv2d(64, 64, kernel_size=3, padding=1)
        self.bn5 = nn.BatchNorm2d(64)
        self.relu5 = nn.ReLU(inplace=True)

        #####

        self.conv_d4 = nn.Conv2d(128, 64, kernel_size=3, padding=1)
        self.bn_d4 = nn.BatchNorm2d(64)
        self.relu_d4 = nn.ReLU(inplace=True)

        self.conv_d3 = nn.Conv2d(128, 64, kernel_size=3, padding=1)
        self.bn_d3 = nn.BatchNorm2d(64)
        self.relu_d3 = nn.ReLU(inplace=True)

        self.conv_d2 = nn.Conv2d(128, 64, kernel_size=3, padding=1)
        self.bn_d2 = nn.BatchNorm2d(64)
        self.relu_d2 = nn.ReLU(inplace=True)

        self.conv_d1 = nn.Conv2d(128, 64, kernel_size=3, padding=1)
        self.bn_d1 = nn.BatchNorm2d(64)
        self.relu_d1 = nn.ReLU(inplace=True)

        self.conv_d0 = nn.Conv2d(64, 1, kernel_size=3, padding=1)

        self.upscore2 = nn.Upsample(
            scale_factor=2, mode="bilinear", align_corners=False
        )
        # self.upscore2 = nn.Upsample(scale_factor=2, mode='bilinear')

    def forward(self, x):
        hx = x
        hx = self.conv0(hx)

        hx1 = self.relu1(self.bn1(self.conv1(hx)))
        hx = self.pool1(hx1)

        hx2 = self.relu2(self.bn2(self.conv2(hx)))
        hx = self.pool2(hx2)

        hx3 = self.relu3(self.bn3(self.conv3(hx)))
        hx = self.pool3(hx3)

        hx4 = self.relu4(self.bn4(self.conv4(hx)))
        hx = self.pool4(hx4)

        hx5 = self.relu5(self.bn5(self.conv5(hx)))

        hx = self.upscore2(hx5)

        d4 = self.relu_d4(self.bn_d4(self.conv_d4(torch.cat((hx, hx4), 1))))
        hx = self.upscore2(d4)

        d3 = self.relu_d3(self.bn_d3(self.conv_d3(torch.cat((hx, hx3), 1))))
        hx = self.upscore2(d3)

        d2 = self.relu_d2(self.bn_d2(self.conv_d2(torch.cat((hx, hx2), 1))))
        hx = self.upscore2(d2)

        d1 = self.relu_d1(self.bn_d1(self.conv_d1(torch.cat((hx, hx1), 1))))

        residual = self.conv_d0(d1)

        return x + residual


def conv3x3(in_planes, out_planes, stride=1) -> nn.Conv2d:
    "3x3 convolution with padding"
    return nn.Conv2d(
        in_planes, out_planes, kernel_size=3, stride=stride, padding=1, bias=False
    )


class BasicBlock(nn.Module):
    expansion: int = 1

    def __init__(self, inplanes: int, planes: int, stride: int = 1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = nn.BatchNorm2d(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = nn.BatchNorm2d(planes)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out += residual
        out = self.relu(out)

        return out


class BASNetModel(PreTrainedModel):
    config_class = BASNetConfig

    def __init__(self, config: BASNetConfig) -> None:
        super().__init__(config)

        resnet = torchvision.models.resnet34(
            weights=torchvision.models.ResNet34_Weights.IMAGENET1K_V1
        )

        ## -------------Encoder--------------

        self.inconv = nn.Conv2d(
            config.n_channels, 64, kernel_size=config.kernel_size, padding=1
        )
        self.inbn = nn.BatchNorm2d(64)
        self.inrelu = nn.ReLU(inplace=True)

        # stage 1
        self.encoder1 = resnet.layer1  # 256
        # stage 2
        self.encoder2 = resnet.layer2  # 128
        # stage 3
        self.encoder3 = resnet.layer3  # 64
        # stage 4
        self.encoder4 = resnet.layer4  # 32

        self.pool4 = nn.MaxPool2d(2, 2, ceil_mode=True)

        # stage 5
        self.resb5_1 = BasicBlock(512, 512)
        self.resb5_2 = BasicBlock(512, 512)
        self.resb5_3 = BasicBlock(512, 512)  # 16

        self.pool5 = nn.MaxPool2d(2, 2, ceil_mode=True)

        # stage 6
        self.resb6_1 = BasicBlock(512, 512)
        self.resb6_2 = BasicBlock(512, 512)
        self.resb6_3 = BasicBlock(512, 512)  # 8

        ## -------------Bridge--------------

        # stage Bridge
        self.convbg_1 = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, dilation=2, padding=2
        )  # 8
        self.bnbg_1 = nn.BatchNorm2d(512)
        self.relubg_1 = nn.ReLU(inplace=True)
        self.convbg_m = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, dilation=2, padding=2
        )
        self.bnbg_m = nn.BatchNorm2d(512)
        self.relubg_m = nn.ReLU(inplace=True)
        self.convbg_2 = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, dilation=2, padding=2
        )
        self.bnbg_2 = nn.BatchNorm2d(512)
        self.relubg_2 = nn.ReLU(inplace=True)

        ## -------------Decoder--------------

        # stage 6d
        self.conv6d_1 = nn.Conv2d(
            1024, 512, kernel_size=config.kernel_size, padding=1
        )  # 16
        self.bn6d_1 = nn.BatchNorm2d(512)
        self.relu6d_1 = nn.ReLU(inplace=True)

        self.conv6d_m = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, dilation=2, padding=2
        )  ###
        self.bn6d_m = nn.BatchNorm2d(512)
        self.relu6d_m = nn.ReLU(inplace=True)

        self.conv6d_2 = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, dilation=2, padding=2
        )
        self.bn6d_2 = nn.BatchNorm2d(512)
        self.relu6d_2 = nn.ReLU(inplace=True)

        # stage 5d
        self.conv5d_1 = nn.Conv2d(
            1024, 512, kernel_size=config.kernel_size, padding=1
        )  # 16
        self.bn5d_1 = nn.BatchNorm2d(512)
        self.relu5d_1 = nn.ReLU(inplace=True)

        self.conv5d_m = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, padding=1
        )  ###
        self.bn5d_m = nn.BatchNorm2d(512)
        self.relu5d_m = nn.ReLU(inplace=True)

        self.conv5d_2 = nn.Conv2d(512, 512, kernel_size=config.kernel_size, padding=1)
        self.bn5d_2 = nn.BatchNorm2d(512)
        self.relu5d_2 = nn.ReLU(inplace=True)

        # stage 4d
        self.conv4d_1 = nn.Conv2d(
            1024, 512, kernel_size=config.kernel_size, padding=1
        )  # 32
        self.bn4d_1 = nn.BatchNorm2d(512)
        self.relu4d_1 = nn.ReLU(inplace=True)

        self.conv4d_m = nn.Conv2d(
            512, 512, kernel_size=config.kernel_size, padding=1
        )  ###
        self.bn4d_m = nn.BatchNorm2d(512)
        self.relu4d_m = nn.ReLU(inplace=True)

        self.conv4d_2 = nn.Conv2d(512, 256, kernel_size=config.kernel_size, padding=1)
        self.bn4d_2 = nn.BatchNorm2d(256)
        self.relu4d_2 = nn.ReLU(inplace=True)

        # stage 3d
        self.conv3d_1 = nn.Conv2d(
            512, 256, kernel_size=config.kernel_size, padding=1
        )  # 64
        self.bn3d_1 = nn.BatchNorm2d(256)
        self.relu3d_1 = nn.ReLU(inplace=True)

        self.conv3d_m = nn.Conv2d(
            256, 256, kernel_size=config.kernel_size, padding=1
        )  ###
        self.bn3d_m = nn.BatchNorm2d(256)
        self.relu3d_m = nn.ReLU(inplace=True)

        self.conv3d_2 = nn.Conv2d(256, 128, kernel_size=config.kernel_size, padding=1)
        self.bn3d_2 = nn.BatchNorm2d(128)
        self.relu3d_2 = nn.ReLU(inplace=True)

        # stage 2d

        self.conv2d_1 = nn.Conv2d(
            256, 128, kernel_size=config.kernel_size, padding=1
        )  # 128
        self.bn2d_1 = nn.BatchNorm2d(128)
        self.relu2d_1 = nn.ReLU(inplace=True)

        self.conv2d_m = nn.Conv2d(
            128, 128, kernel_size=config.kernel_size, padding=1
        )  ###
        self.bn2d_m = nn.BatchNorm2d(128)
        self.relu2d_m = nn.ReLU(inplace=True)

        self.conv2d_2 = nn.Conv2d(128, 64, kernel_size=config.kernel_size, padding=1)
        self.bn2d_2 = nn.BatchNorm2d(64)
        self.relu2d_2 = nn.ReLU(inplace=True)

        # stage 1d
        self.conv1d_1 = nn.Conv2d(
            128, 64, kernel_size=config.kernel_size, padding=1
        )  # 256
        self.bn1d_1 = nn.BatchNorm2d(64)
        self.relu1d_1 = nn.ReLU(inplace=True)

        self.conv1d_m = nn.Conv2d(
            64, 64, kernel_size=config.kernel_size, padding=1
        )  ###
        self.bn1d_m = nn.BatchNorm2d(64)
        self.relu1d_m = nn.ReLU(inplace=True)

        self.conv1d_2 = nn.Conv2d(64, 64, kernel_size=config.kernel_size, padding=1)
        self.bn1d_2 = nn.BatchNorm2d(64)
        self.relu1d_2 = nn.ReLU(inplace=True)

        ## -------------Bilinear Upsampling--------------
        self.upscore6 = nn.Upsample(
            scale_factor=32, mode="bilinear", align_corners=False
        )  ###
        self.upscore5 = nn.Upsample(
            scale_factor=16, mode="bilinear", align_corners=False
        )
        self.upscore4 = nn.Upsample(
            scale_factor=8, mode="bilinear", align_corners=False
        )
        self.upscore3 = nn.Upsample(
            scale_factor=4, mode="bilinear", align_corners=False
        )
        self.upscore2 = nn.Upsample(
            scale_factor=2, mode="bilinear", align_corners=False
        )

        # self.upscore6 = nn.Upsample(scale_factor=32, mode='bilinear') ###
        # self.upscore5 = nn.Upsample(scale_factor=16, mode='bilinear')
        # self.upscore4 = nn.Upsample(scale_factor=8, mode='bilinear')
        # self.upscore3 = nn.Upsample(scale_factor=4, mode='bilinear')
        # self.upscore2 = nn.Upsample(scale_factor=2, mode='bilinear')

        ## -------------Side Output--------------
        self.outconvb = nn.Conv2d(512, 1, kernel_size=3, padding=1)
        self.outconv6 = nn.Conv2d(512, 1, kernel_size=3, padding=1)
        self.outconv5 = nn.Conv2d(512, 1, kernel_size=3, padding=1)
        self.outconv4 = nn.Conv2d(256, 1, kernel_size=3, padding=1)
        self.outconv3 = nn.Conv2d(128, 1, kernel_size=3, padding=1)
        self.outconv2 = nn.Conv2d(64, 1, kernel_size=3, padding=1)
        self.outconv1 = nn.Conv2d(64, 1, kernel_size=3, padding=1)

        ## -------------Refine Module-------------
        self.refunet = RefUnet(1, 64)

        self.post_init()

    def forward(
        self, pixel_values: torch.Tensor, return_dict: Optional[bool] = None
    ) -> Union[Tuple, BASNetModelOutput]:
        hx = pixel_values

        ## -------------Encoder-------------
        hx = self.inconv(hx)
        hx = self.inbn(hx)
        hx = self.inrelu(hx)

        h1 = self.encoder1(hx)  # 256
        h2 = self.encoder2(h1)  # 128
        h3 = self.encoder3(h2)  # 64
        h4 = self.encoder4(h3)  # 32

        hx = self.pool4(h4)  # 16

        hx = self.resb5_1(hx)
        hx = self.resb5_2(hx)
        h5 = self.resb5_3(hx)

        hx = self.pool5(h5)  # 8

        hx = self.resb6_1(hx)
        hx = self.resb6_2(hx)
        h6 = self.resb6_3(hx)

        ## -------------Bridge-------------
        hx = self.relubg_1(self.bnbg_1(self.convbg_1(h6)))  # 8
        hx = self.relubg_m(self.bnbg_m(self.convbg_m(hx)))
        hbg = self.relubg_2(self.bnbg_2(self.convbg_2(hx)))

        ## -------------Decoder-------------

        hx = self.relu6d_1(self.bn6d_1(self.conv6d_1(torch.cat((hbg, h6), 1))))
        hx = self.relu6d_m(self.bn6d_m(self.conv6d_m(hx)))
        hd6 = self.relu6d_2(self.bn5d_2(self.conv6d_2(hx)))

        hx = self.upscore2(hd6)  # 8 -> 16

        hx = self.relu5d_1(self.bn5d_1(self.conv5d_1(torch.cat((hx, h5), 1))))
        hx = self.relu5d_m(self.bn5d_m(self.conv5d_m(hx)))
        hd5 = self.relu5d_2(self.bn5d_2(self.conv5d_2(hx)))

        hx = self.upscore2(hd5)  # 16 -> 32

        hx = self.relu4d_1(self.bn4d_1(self.conv4d_1(torch.cat((hx, h4), 1))))
        hx = self.relu4d_m(self.bn4d_m(self.conv4d_m(hx)))
        hd4 = self.relu4d_2(self.bn4d_2(self.conv4d_2(hx)))

        hx = self.upscore2(hd4)  # 32 -> 64

        hx = self.relu3d_1(self.bn3d_1(self.conv3d_1(torch.cat((hx, h3), 1))))
        hx = self.relu3d_m(self.bn3d_m(self.conv3d_m(hx)))
        hd3 = self.relu3d_2(self.bn3d_2(self.conv3d_2(hx)))

        hx = self.upscore2(hd3)  # 64 -> 128

        hx = self.relu2d_1(self.bn2d_1(self.conv2d_1(torch.cat((hx, h2), 1))))
        hx = self.relu2d_m(self.bn2d_m(self.conv2d_m(hx)))
        hd2 = self.relu2d_2(self.bn2d_2(self.conv2d_2(hx)))

        hx = self.upscore2(hd2)  # 128 -> 256

        hx = self.relu1d_1(self.bn1d_1(self.conv1d_1(torch.cat((hx, h1), 1))))
        hx = self.relu1d_m(self.bn1d_m(self.conv1d_m(hx)))
        hd1 = self.relu1d_2(self.bn1d_2(self.conv1d_2(hx)))

        ## -------------Side Output-------------
        db = self.outconvb(hbg)
        db = self.upscore6(db)  # 8->256

        d6 = self.outconv6(hd6)
        d6 = self.upscore6(d6)  # 8->256

        d5 = self.outconv5(hd5)
        d5 = self.upscore5(d5)  # 16->256

        d4 = self.outconv4(hd4)
        d4 = self.upscore4(d4)  # 32->256

        d3 = self.outconv3(hd3)
        d3 = self.upscore3(d3)  # 64->256

        d2 = self.outconv2(hd2)
        d2 = self.upscore2(d2)  # 128->256

        d1 = self.outconv1(hd1)  # 256

        ## -------------Refine Module-------------
        dout = self.refunet(d1)  # 256

        dout_act = torch.sigmoid(dout)
        d1_act = torch.sigmoid(d1)
        d2_act = torch.sigmoid(d2)
        d3_act = torch.sigmoid(d3)
        d4_act = torch.sigmoid(d4)
        d5_act = torch.sigmoid(d5)
        d6_act = torch.sigmoid(d6)
        db_act = torch.sigmoid(db)

        side_outputs = (
            dout_act,
            d1_act,
            d2_act,
            d3_act,
            d4_act,
            d5_act,
            d6_act,
            db_act,
        )
        if not return_dict:
            return (side_outputs,)

        return BASNetModelOutput(
            activated=BasNetSideOutput(*side_outputs),
        )


def convert_from_checkpoint(
    repo_id: str, filename: str, config: Optional[BASNetConfig] = None
) -> BASNetModel:
    from huggingface_hub import hf_hub_download

    checkpoint_path = hf_hub_download(repo_id=repo_id, filename=filename)

    config = config or BASNetConfig()
    model = BASNetModel(config)

    logger.info(f"Loading checkpoint from {checkpoint_path}")
    state_dict = torch.load(checkpoint_path)

    model.load_state_dict(state_dict, strict=True)
    model.eval()

    return model