CV-Transformer集合

网络名称	机构	领域	最低指标	最高指标	原文	代码	博文	期刊	时间
Transformer	谷歌	NLP	——	——	[Link]	[TensorFlow]	[Link]	NIPS2017	2017.07
ViT	谷歌	CV	——	——	[Link]	[TensorFlow]	[Link]	-	2020.10
DeiT	脸书	CV	72.2/5M/	85.2/87M/	[Link]	[PyTorch]	[Link]	-	2020.12
BoT	UC Berkeley	CV	77.0/20.8M/4.27G	84.7/75.1M/45.8B	[Link]	[PyTorch]	[Link]	-	2021.01
CaiT	微软	CV	83.3/68M/13.9B	86.5/356M/329.6B	[Link]	[PyTorch]	[Link]	-	2021.03
CeiT	商汤	CV	76.4/6.4M/1.2G	83.3/24.2M/12.9G	[Link]	[PyTorch]	[Link]	-	2021.03
PiT	NAVER AI LAB	CV	74.6/4.9M/0.7B	84.0/73.8M/12.5B	[Link]	[PyTorch]	[Link]	-	2021.03
TNT	诺亚	CV	81.5/23.8M/5.2B	82.9/65.6M/14.1B	[Link]	[PyTorch]	[Link]	-	2021.03
Twins	美团	CV	81.2/24.1M/3.8G	83.7/99.2M/15.1G	[Link]	[PyTorch]	[Link]	-	2021.04
SwinT	微软	CV	81.3/29M/4.5G	84.2/88M/47G	[Link]	[PyTorch]	[Link]	-	2021.05
ResT	南大	CV	77.2/10.49M/1.4G	83.6/51.63M/7.9G	[Link]	[PyTorch]	[Link]	-	2021.05
NesT	谷歌	CV	81.5/17M/5.8G	83.8/68M/17.9G	[Link]	[TensorFlow]	[Link]	-	2021.05
Conformer	中科大	CV	81.3/23.5M/5.2G	84.1/83.3M/23.3G	[Link]	[PyTorch]	[Link]	-	2021.05
CvT	McGill Univercity	CV	81.6/20M/4.5G	83.3/32M/24.9G	[Link]	[PyTorch]	[Link]	-	2021.05
Refiner	新加坡国立大学	CV	83.6/25M/7.2B	85.9/81M/98G	[Link]	[PyTorch]	[Link]	-	2021.06
Focal	脸书	CV	82.2/29.1M/4.9G	83.8/89.8M/16.0G	[Link]	[PyTorch]	[Link]	-	2021.07

❗❗❗指标指在ImageNet1k数据集上的Top1精度AP/参数量/计算量。
⚡⚡Transformer代码集合[Pytorch_Transformer_Models]⚡⚡