论文阅读清单

:maxdepth: 2

神经网络基础(basis)

Num	Title	Field	Desc	Author	Time	read
1	ADAM: A METHOD FOR STOCHASTIC OPTIMIZATION				2015
2	Wide & Deep Learning for Recommender Systems				2016
3	Targeted Dropout

批量&正则化(batch&normalization)

Num	Title	Field	Desc	Author	Time	read
1	Batch Normalization: Accelerating Deep Network Training b y Reducing Internal Covariate Shift		批量正则化论文		2015
2	Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models		ReNorm算法论文		2017
	Instance Normalization: The Missing Ingredient for Fast Stylization		实例归一化论文		2017
3	Group Normalization		GroupNorm算法论文		2018
4	DIFFERENTIABLE LEARNING-TO-NORMALIZE VIA SWITCHABLE NORMALIZATION		SwitchableNorm算法论文		2019

注意力部分(attention)

Num	Title	Field	Desc	Author	Time	read
1	Attention-Based Models for Speech Recognition		混合注意力机制论文		2015
2	Effective Approaches to Attention-based Neural Machine Translation		孪生注意力论文		2015
3	Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks		各自升级的孪生注意力论文		2016
4	NEURAL MACHINE TRANSLATION BY JOINTLY LEARNING TO ALIGN AND TRANSLATE		孪生注意力论文		2016
5	Attention Is All You Need		大道至简的注意力论文		2017
6	Online and Linear-Time Attention by Enforcing Monotonic Alignments		单调注意力机制论文		2017

高级卷积网络知识(Convolutional)

Num	Title	Field	Desc	Author	Time	read
1	Convolutional Neural Networks for Sentence Classification		卷积网络新玩法TextCNN模型		2014
2	MATRIX CAPSULES WITH EM ROUTING		矩阵胶囊网络与EM路由算法
3	Dynamic Routing Between Capsules		胶囊网络与动态路由的论文		2017
4	Information Aggregation via Dynamic Routing for Sequence Encoding		胶囊网络的其它用处		2018

循环神经网络(RNN)

Num	Title	Field	Desc	Author	Time	read
1	QUASI-RECURRENT NEURAL NETWORKS		QRNN		2016
2	Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN		IndRNN		2018
3	THE UNREASONABLE EFFECTIVENESS OF THE FORGET GATE		IndRNN		2018
4	Simple Recurrent Units for Highly Parallelizable Recurrence		SRU		2018
5

Transformer

Num	Title	Field	Desc	Author	Time	read

AI合成部分(GAN)

Num	Title	Field	Desc	Author	Time	read
1	Improved Training of Wasserstein GANs		RNN.WGAN		2017
2	TACOTRON: TOWARDS END-TO-END SPEECH SYNTHESIS		Tacotron与Tacotron-2		2017
4	AttGAN: Facial Attribute Editing by Only Changing What You Want		AttGAN		2018
5	DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial Networks		DeblurGAN		2018
6	NATURAL TTS SYNTHESIS BY CONDITIONING WAVENET ON MEL SPECTROGRAM PREDICTIONS		Tacotron&Tacotron-2		2018

目标分割(SEG)

Num	Title	Field	Desc	Author	Time	read
7	Fully Convolutional Networks for Semantic Segmentation	目标分割	FCN
8	U-Net：Convolutional Networks for Biomedical	目标分割	U-Net
9	Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs	目标分割	Deeplabv1
10	Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs	目标分割	Deeplabv2
11	Rethinking Atrous Convolution for Semantic Image Segmentation	目标分割	Deeplabv3
12	Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation	目标分割	Deeplabv3+
13	Mask R-CNN	目标分割	Mask R-CNN
14	Feature Pyramid Networks for Object Detection	目标分割	FPN
15	Focal Loss for Dense Object Detection	目标分割	RetinaNet

---

目标检测(OBJ)

Num	Title	Field	Desc	Author	Time	read
1	Rich feature hierarchies for accurate object detection and semantic segmentation	目标检测	R-CNN
2	Fast R-CNN	目标检测	Fast R-CNN
3	Faster R-CNN：Towards Real-Time Object	目标检测	Faster R-CNN
4	Mask R-CNN	目标检测	Mask R-CNN
5	SSD:Single Shot MultiBox Detector	目标检测	SSD
6	Feature Pyramid Networks for Object Detection	目标分割	FPN
7	Focal Loss for Dense Object Detection	目标分割	RetinaNet
8	Bag of Freebies for Training Object Detection Neural Networks	目标分割
9	You Only Look One-Unified, Real-Time Object Detection	目标分割	YOLOv1
10	YOLO9000：Better, Faster, Stronger	目标分割	YOLOv2
11	YOLOv3：An Incremental Improvement	目标分割	YOLOv3
12	YOLOv4：Optimal Speed and Accuracy of Object Detection	目标分割	YOLOv4
13	PP-YOLO：An Effective and Efficient Implementation of Object Detector	目标分割	PP-YOLO
14	PP-YOLOv2：A Practical Object Detector	目标分割	PP-YOLO2

---

图像分类(CLAS)

Num	Title	Field	Desc	Author	Time	read
1	Gradient-based Learning Applied to Document Recognition		LeNet
2	ImageNet Classification with Deep Convolutional		AlexNet
3	Visualizing and Understanding Convolutional Networks		ZFNet
4	VERY DEEP CONVOLUTIONAL		VGG
5	Going deeper with convolutions		GoogleNet,Inceptionv1
6	Batch Normalization-Accelerating Deep Network Training b
7	Rethinking the Inception Architecture for Computer Vision		Inceptionv3
8	Inception-v4：Inception-ResNet and the Impact of Residual Connections on Learning		Inception-v4
9	Xception：Deep Learning with Depthwise Separable Convolutions		Xception
10	Deep Residual Learning for Image Recognition		ResNet
11	Aggregated Residual Transformations for Deep Neural Networks		ResNeXt
12	Densely Connected Convolutional Networks		DenseNet
13	Learning Transferable Architectures for Scalable Image Recognition		NASNet-A
14	MobileNets-Efficient Convolutional Neural Networks for Mobile Vision		SENet
15	MobileNets- Efficient Convolutional Neural Networks for Mobile Vision		MobileNets-v1
16	MobileNetV2：Inverted Residuals and Linear Bottlenecks		MobileNets-v2
17	Searching for MobileNetV3		MobileNets-v3
18	ShuffleNet：An Extremely Efficient Convolutional Neural Network for Mobile		ShuffleNet
19	ShuffleNet V2：Practical Guidelines for Efficient		ShuffleNet-v2
20	Bag of Tricks for Image Classification with Convolutional Neural Networks
21	EfficientNet：Rethinking Model Scaling for Convolutional Neural Networks		EfficientNet
22	EfficientNetV2：Smaller Models and Faster Training		EfficientNet-v2
23	CSPNET-A NEW BACKBONE THAT CAN ENHANCE LEARNING		CSPNET-A
24	High-Performance Large-Scale Image Recognition Without Normalization		NFNets
25	AN IMAGE IS WORTH 16X16 WORDS-T RANSFORMERS FOR I MAGE R ECOGNITION AT S CALE		Vision Transformer
26	Training data-efficient image transformers		DeiT
27	Swin Transformer-Hierarchical Vision Transformer using Shifted Windows		Swin Transformer

---

自然语言处理(NLP)

Num	Title	Field	Desc	Author	Time	read
1	Attention Is All You Need	注意力机制	Attention

---

多模态(MultiModal Learning)

Num	Title	Field	Desc	Author	Time	read
					2022
	BLIP: Bootstrapping Language-Image Pre-training		视觉语言预训练	Introduced by Li et al.	2022
	BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models		使用冻结图像编码器和大型语言模型进行引导语言图像预训练	Junnan Li，Dongxu Li，Silvio Savarese，Steven Hoi	2023

大语言模型(Large Language Models)

Num	Title	Field	Desc	Author	Time	read
	OPT: OPT : Open Pre-trained Transformer Language Models		开放预训练的 Transformer 语言模型	Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen, Christopher Dewan, Mona Diab, Xian Li, Xi Victoria Lin, Todor Mihaylov, Myle Ott, Sam Shleifer, Kurt Shuster, Daniel Simig, Punit Singh Koura, Anjali Sridhar, Tianlu Wang, Luke Zettlemoyer
	GPT-v1:Improving Language Understanding by Generative Pre-Training	GPT&LLM
	GPT-v2:Language Models are Unsupervised Multitask Learners	GPT&LLM
	GPT-v3:Language Models are Few-Shot Learners	GPT&LLM
	GPT-v4:GPT-4 Technical Report	GPT&LLM