（CCNET）criss-cross attention network学习笔记

原文：https://arxiv.org/pdf/1811.11721.pdf

记录一下，目前习惯用笔写的：

cc attention这个模块的代码：

from .functions import CrissCrossAttention'''
This code is borrowed from Serge-weihao/CCNet-Pure-Pytorch
'''import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.nn import Softmaxdef INF(B,H,W):return -torch.diag(torch.tensor(float("inf")).cuda().repeat(H),0).unsqueeze(0).repeat(B*W,1,1)class CrissCrossAttention(nn.Module):""" Criss-Cross Attention Module"""def __init__(self, in_dim):super(CrissCrossAttention,self).__init__()self.query_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim//8, kernel_size=1)self.key_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim//8, kernel_size=1)self.value_conv = nn.Conv2d(in_channels=in_dim, out_channels=in_dim, kernel_size=1)self.softmax = Softmax(dim=3)self.INF = INFself.gamma = nn.Parameter(torch.zeros(1))def forward(self, x):m_batchsize, _, height, width = x.size()proj_query = self.query_conv(x)proj_query_H = proj_query.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height).permute(0, 2, 1)proj_query_W = proj_query.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width).permute(0, 2, 1)proj_key = self.key_conv(x)proj_key_H = proj_key.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height)proj_key_W = proj_key.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width)proj_value = self.value_conv(x)proj_value_H = proj_value.permute(0,3,1,2).contiguous().view(m_batchsize*width,-1,height)proj_value_W = proj_value.permute(0,2,1,3).contiguous().view(m_batchsize*height,-1,width)energy_H = (torch.bmm(proj_query_H, proj_key_H)+self.INF(m_batchsize, height, width)).view(m_batchsize,width,height,height).permute(0,2,1,3)energy_W = torch.bmm(proj_query_W, proj_key_W).view(m_batchsize,height,width,width)concate = self.softmax(torch.cat([energy_H, energy_W], 3))att_H = concate[:,:,:,0:height].permute(0,2,1,3).contiguous().view(m_batchsize*width,height,height)#print(concate)#print(att_H) att_W = concate[:,:,:,height:height+width].contiguous().view(m_batchsize*height,width,width)out_H = torch.bmm(proj_value_H, att_H.permute(0, 2, 1)).view(m_batchsize,width,-1,height).permute(0,2,3,1)out_W = torch.bmm(proj_value_W, att_W.permute(0, 2, 1)).view(m_batchsize,height,-1,width).permute(0,2,1,3)#print(out_H.size(),out_W.size())return self.gamma*(out_H + out_W) + xif __name__ == '__main__':model = CrissCrossAttention(64)x = torch.randn(2, 64, 5, 6)out = model(x)print(out.shape)

（CCNET）criss-cross attention network学习笔记相关推荐

深度学习（二十六）Network In Network学习笔记-ICLR 2014
Network In Network学习笔记原文地址:http://blog.csdn.net/hjimce/article/details/50458190 作者:hjimce 一.相关理论本篇 ...
深度学习（二十六）Network In Network学习笔记
Network In Network学习笔记原文地址:http://blog.csdn.net/hjimce/article/details/50458190 作者:hjimce 一.相关理论本篇 ...
NIN(Network in Network)学习笔记
NIN(Network in Network)学习笔记一.前言 <Network In Network>是一篇比较老的文章了(2014年ICLR的一篇paper),是当时比较牛逼的一篇论 ...
matlab rbm 语音,Deep Belief Network 学习笔记-RBM
Deep Belief Network 学习笔记-RBM By Placebo (纯属个人笔记) 第一次知道deep learning,是上学期dengli博士来实验室的一次报告,他讲到,当神经网络的 ...
GNN金融应用之Classifying and Understanding Financial Data Using Graph Neural Network学习笔记
Classifying and Understanding Financial Data Using Graph Neural Network 摘要 1. 概述 2. 数据表示-加权图 3. GNN利 ...
图神经网络框架DGL实现Graph Attention Network (GAT)笔记
参考列表: [1]深入理解图注意力机制 [2]DGL官方学习教程一 --基础操作&消息传递 [3]Cora数据集介绍+python读取一.DGL实现GAT分类机器学习论文程序摘自[1],该 ...
network 学习笔记
#cat /etc/modprobe.conf Ethernet : eth0,eth1,ethN Token Ring: tr0,tr1,trN FDDI : fddi0,fddi1,fddiN ( ...
Neural Network学习笔记2
torch.nn: Containers: 神经网络骨架 Convolution Layers 卷积层 Pooling Layers 池化层 Normalization Layers 正则化层 No ...
内容分发网络 - Content Delivery Network 学习笔记
缓存是将文件副本存储在缓存或临时存储位置的过程,以便可以更快地访问它们.从技术上讲,缓存是文件或数据副本的任何临时存储位置,但该术语通常用于指代 Internet 技术. Web 浏览器缓存 HTML ...
Focusing Attention Network（FAN）自然图像文本识别学习笔记
Focusing Attention: Towards Accurate Text Recognition in Natural Images Author: Zhanzhan Cheng,Fan B ...

（CCNET）criss-cross attention network学习笔记

（CCNET）criss-cross attention network学习笔记相关推荐

最新文章

热门文章