Bottleneck Transformer
时间: 2023-10-14 20:04:33 浏览: 51
Bottleneck Transformer(简称BotNet)是一种深度学习模型,通过在Multi-Head Self-Attention结构前后加上1x1卷积,从而扩展了传统的Transformer结构。BotNet和ViT中的Transformer block有一定的相似性,可以将具有MHSA的ResNet bottleneck块视为具有bottleneck结构的Transformer块,尽管它们在残差连接、标准化层等方面存在一些微小差异。这个模型在2021年由Google的研究人员在论文"Bottleneck Transformers for Visual Recognition"中提出,并引入了CNN与Attention的融合网络。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* *3* [深度学习之图像分类(十九)-- Bottleneck Transformer(BoTNet)网络详解](https://blog.csdn.net/baidu_36913330/article/details/120218954)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 50%"]
- *2* [第53步 深度学习图像识别:Bottleneck Transformer建模(Pytorch)](https://blog.csdn.net/qq_30452897/article/details/131741943)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 50%"]
[ .reference_list ]