data/lib/models/modules/base_oc_block.py

##+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
## Created by: RainbowSecret
## Modified from: https://github.com/AlexHex7/Non-local_pytorch
## Microsoft Research
## yuyua@microsoft.com
## Copyright (c) 2018
##
## Ocnet: Object context network for scene parsing
##
## This source code is licensed under the MIT-style license found in the
## LICENSE file in the root directory of this source tree 
##+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

import os
import sys
import pdb
import torch
from torch import nn
from torch.nn import functional as F

from lib.models.tools.module_helper import ModuleHelper


class _SelfAttentionBlock(nn.Module):
    '''
    The basic implementation for self-attention block/non-local block
    Input:
        N X C X H X W
    Parameters:
        in_channels       : the dimension of the input feature map
        key_channels      : the dimension after the key/query transform
        value_channels    : the dimension after the value transform
        scale             : choose the scale to downsample the input feature maps (save memory cost)
    Return:
        N X C X H X W
        position-aware context features.(w/o concate or add with the input)
    '''
    def __init__(self, 
                 in_channels, 
                 key_channels, 
                 value_channels, 
                 out_channels=None, 
                 scale=1, 
                 bn_type=None):
        super(_SelfAttentionBlock, self).__init__()
        self.scale = scale
        self.in_channels = in_channels
        self.out_channels = out_channels
        self.key_channels = key_channels
        self.value_channels = value_channels
        if out_channels == None:
            self.out_channels = in_channels
        self.pool = nn.MaxPool2d(kernel_size=(scale, scale))
        self.f_key = nn.Sequential(
            nn.Conv2d(in_channels=self.in_channels, out_channels=self.key_channels,
                kernel_size=1, stride=1, padding=0),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
            nn.Conv2d(in_channels=self.key_channels, out_channels=self.key_channels,
                kernel_size=1, stride=1, padding=0),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
        )
        self.f_query = nn.Sequential(
            nn.Conv2d(in_channels=self.in_channels, out_channels=self.key_channels,
                kernel_size=1, stride=1, padding=0),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
            nn.Conv2d(in_channels=self.key_channels, out_channels=self.key_channels,
                kernel_size=1, stride=1, padding=0),
            ModuleHelper.BNReLU(self.key_channels, bn_type=bn_type),
        )

        self.f_value = nn.Conv2d(in_channels=self.in_channels, out_channels=self.value_channels,
            kernel_size=1, stride=1, padding=0)
        self.W = nn.Conv2d(in_channels=self.value_channels, out_channels=self.out_channels,
            kernel_size=1, stride=1, padding=0)
        nn.init.constant_(self.W.weight, 0)
        nn.init.constant_(self.W.bias, 0)

    def forward(self, x):
        batch_size, h, w = x.size(0), x.size(2), x.size(3)
        if self.scale > 1:
            x = self.pool(x)   
        value = self.f_value(x).view(batch_size, self.value_channels, -1)
        value = value.permute(0, 2, 1)
        query = self.f_query(x).view(batch_size, self.key_channels, -1)
        query = query.permute(0, 2, 1)
        key = self.f_key(x).view(batch_size, self.key_channels, -1)

        sim_map = torch.matmul(query, key)
        sim_map = (self.key_channels**-.5) * sim_map
        sim_map = F.softmax(sim_map, dim=-1)

        context = torch.matmul(sim_map, value)
        context = context.permute(0, 2, 1).contiguous()
        context = context.view(batch_size, self.value_channels, *x.size()[2:])
        context = self.W(context)
        if self.scale > 1:
            context = F.interpolate(input=context, size=(h, w), mode='bilinear', align_corners=True)
        return context


class SelfAttentionBlock2D(_SelfAttentionBlock):
    def __init__(self, 
                 in_channels, 
                 key_channels, 
                 value_channels, 
                 out_channels=None, 
                 scale=1, 
                 bn_type=None):
        super(SelfAttentionBlock2D, self).__init__(in_channels,
                                                    key_channels,
                                                    value_channels,
                                                    out_channels,
                                                    scale, bn_type)


class BaseOC_Module(nn.Module):
    """
    Implementation of the BaseOC module
    Parameters:
        in_features / out_features: the channels of the input / output feature maps.
        dropout: we choose 0.05 as the default value.
        size: you can apply multiple sizes. Here we only use one size.
    Return:
        features fused with Object context information.
    """
    def __init__(self, 
                 in_channels, 
                 out_channels, 
                 key_channels, 
                 value_channels, 
                 dropout, 
                 sizes=([1]), 
                 bn_type=None):
        super(BaseOC_Module, self).__init__()
        self.stages = []
        self.stages = nn.ModuleList([self._make_stage(in_channels, in_channels,
                                                      key_channels, value_channels, size, bn_type) for size in sizes])
        self.conv_bn_dropout = nn.Sequential(
            nn.Conv2d(2*in_channels, out_channels, kernel_size=1, padding=0),
            ModuleHelper.BNReLU(out_channels, bn_type=bn_type),
            nn.Dropout2d(dropout)
        )

    def _make_stage(self, in_channels, output_channels, key_channels, value_channels, size, bn_type):
        return SelfAttentionBlock2D(in_channels,
                                    key_channels,
                                    value_channels,
                                    output_channels, 
                                    size, 
                                    bn_type=bn_type)
        
    def forward(self, feats):
        priors = [stage(feats) for stage in self.stages]
        context = priors[0]
        for i in range(1, len(priors)):
            context += priors[i]
        output = self.conv_bn_dropout(torch.cat([context, feats], 1))
        return output


class BaseOC_Context_Module(nn.Module):
    """
    Output only the context features.
    Parameters:
        in_features / out_features: the channels of the input / output feature maps.
        dropout: specify the dropout ratio
        fusion: We provide two different fusion method, "concat" or "add"
        size: we find that directly learn the attention weights on even 1/8 feature maps is hard.
    Return:
        features after "concat" or "add"
    """
    def __init__(self, in_channels, out_channels, key_channels, value_channels, dropout=0, sizes=([1]), bn_type=None):
        super(BaseOC_Context_Module, self).__init__()
        self.stages = []
        self.stages = nn.ModuleList([self._make_stage(in_channels, out_channels,
                                                      key_channels, value_channels, size, bn_type) for size in sizes])
        self.conv_bn_dropout = nn.Sequential(
            ModuleHelper.BNReLU(out_channels, bn_type=bn_type),
            nn.Dropout2d(dropout),
            )

    def _make_stage(self, in_channels, output_channels, key_channels, value_channels, size, bn_type):
        return SelfAttentionBlock2D(in_channels,
                                    key_channels,
                                    value_channels,
                                    output_channels, 
                                    size, bn_type=bn_type)
        
    def forward(self, feats):
        priors = [stage(feats) for stage in self.stages]
        context = priors[0]
        for i in range(1, len(priors)):
            context += priors[i]
        output = self.conv_bn_dropout(context)
        return output

if __name__ == "__main__":
    os.environ["CUDA_VISIBLE_DEVICES"] = '0'
    feats = torch.randn((1, 2048, 128, 128)).cuda()

    conv_3x3 = nn.Sequential(
        nn.Conv2d(2048, 512, kernel_size=3, stride=1, padding=1),
        ModuleHelper.BNReLU(512, bn_type='torchsyncbn'),
    )
    baseoc_infer = BaseOC_Module(in_channels=512,
                                 out_channels=512, 
                                 key_channels=256,
                                 value_channels=256, 
                                 sizes=([1]),
                                 dropout=0, 
                                 bn_type='torchsyncbn')
    baseoc_infer.eval()
    conv_3x3.eval()
    baseoc_infer.cuda()
    conv_3x3.cuda()
    
    def count_parameters(model):
        return sum(p.numel() for p in model.parameters() if p.requires_grad)

    avg_time = 0
    avg_mem  = 0
    import time
    with torch.no_grad():
        for i in range(100):
            start_time = time.time()
            outputs = conv_3x3(feats)
            outputs = baseoc_infer(outputs)
            torch.cuda.synchronize()
            avg_time += (time.time() - start_time)
            avg_mem  += (torch.cuda.max_memory_allocated()-feats.element_size() * feats.nelement())

    print("Average Parameters : {}".format(count_parameters(baseoc_infer)+count_parameters(conv_3x3)))
    print("Average Running Time: {}".format(avg_time/100))
    print("Average GPU Memory: {:.2f} MB".format(avg_mem / 100 / 2**20))