迁移学习 材料集合

最新推荐文章于 2021-03-11 19:00:27 发布
最新推荐文章于 2021-03-11 19:00:27 发布
阅读量650 收藏 11
点赞数 1
分类专栏： MyDLNote-Unsupervised domain adaptation 文章标签： 机器学习 深度学习
原文链接：https://github.com/yuntaodu/Transfer-learning-materials
版权

迁移学习 材料集合

目录

迁移学习 材料集合

Book

novel_papers

1) novel_papers on transfer learning

2) novel_papers on related fileds

更多 DA awesome​​​​​​​

入门参考

小结

Excellent Scholars

新论文追踪

科研方法论

Presentation

大部分内容 转自 GitHub：https://github.com/yuntaodu/Transfer-learning-materials

Book

《迁移学习简明手册》

https://github.com/jindongwang/transferlearning-tutorial

 

novel_papers

1) novel_papers on transfer learning

numberTitleConference/journel + yearCodeKeywordsBenenit for us
54Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID (paper)NIPS 2020codecontrastive learning, DA, Re-IDcontrastive learning + DA
53Measuring Information Transfer in Neural Networks (paper)arvix 2020  maybe useful for DA
52Open-Set Hypothesis Transfer with Semantic Consistency (paper)arvix 2020 source free, open set 
51Learning the Stein Discrepancy for Training and Evaluating Energy-Based Models without Sampling (paper)ICML 2020 stein discrepancya new metric that is never used in DA
50Impact of ImageNet Model Selection on Domain Adaptation(paper)WACV 2020 workshop shallow methods with different deep features实验结果很迷惑
49Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks(paper)arvix 2020 pretraininggood papers
48Attract, Perturb, and Explore: Learning a Feature Alignment Network for Semi-supervised Domain Adaptation (paper)ECCV 2020codeSSDA, intar-domain discrepancygood questions
47Measuring Information Transfer in Neural Networks(paper)   interesting paper
46Neural transfer learning for natural language processing（paper)2019 PDH thesis NLP, transfer lerningvery detailed related work
45When Semi-Supervised Learning Meets Transfer Learning: Training Strategies, Models and Datasets(paper)  SSL, TL, experimentsmany results related to multiple SSL methods can be seen in this paper
44Unsupervised Transfer Learning for Spatiotemporal Predictive Networks (paper)ICML 2020   
43Estimating Generalization under Distribution Shifts via Domain-Invariant Representations (paper)ICML 2020codenew theoryrecommend to read
42Implicit Class-Conditioned Domain Alignment for Unsupervised Domain Adaptation (paper)ICML 2020codeideas from theoryrecommend to read
41LEEP: A New Measure to Evaluate Transferability of Learned Representations (paper)ICML 2020 new metric for transferabilityeasy to use for other tasks
40Label-Noise Robust Domain AdaptationICML2020  the author is a rising star
39Progressive Graph Learning for Open-Set Domain Adaptation (paper)ICML 2020codeopen set DA 
38Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation (paper)ICML 2020codesource-free DArecommend to read, new trneds
37Graph Optimal Transport for Cross-Domain Alignment (paper)ICML 2020 graph for DAconnenction with GCN
36Learning Deep Kernels for Non-Parametric Two-Sample Tests (paper)ICML 2020codeextend MMD to deep 
35Adversarial-Learned Loss for Domain AdaptationAAAI 2020 noisy label, adversarial learning 
34Transfer Learning for Anomaly Detection through Localized and Unsupervised Instance SelectionAAAI 2020 transfer learning, anamaly detection 
33Dynamic Instance Normalization for Arbitrary Style TransferAAAI 2020 dynamic instance normalization 
32AdaFilter: Adaptive Filter Fine-Tuning for Deep Transfer LearningAAAI 2020 gated output, fine-tune 
31Bi-Directional Generation for Unsupervised Domain AdaptationAAAI 2020 differert feature extractor, different classifiersconnection with ICML 2019, the third term
30Discriminative Adversarial Domain AdaptationAAAI 2020 discriminative information with adversarial learning 
29Domain Generalization Using a Mixture of Multiple Latent DomainsAAAI 2020   
28Multi-Source Distilling Domain AdaptationAAAI 2020 multi-source 
27Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-SupervisionCVPR 2020codeEntropy adversarial based 
26Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition from a Domain Adaptation PerspectiveCVPR 2020 long-tailed 
25Unsupervised Domain Adaptation via Structurally Regularized Deep ClusteringCVPR 2020codecluster 
24Stochastic Classifiers for Unsupervised Domain AdaptationCVPR 2020 stochastic two classifierssimialer to MCD
23Progressive Adversarial Networks for Fine-Grained Domain AdaptationCVPR 2020 fine-grainedsimilar to mutil-aspect opinion analysis
22Model Adaptation: Unsupervised Domain Adaptation without Source DataCVPR 2020  Recommend to read, new problems
21Towards Inheritable Models for Open-Set Domain AdaptationCVPR 2020code  
20Joint Disentangling and Adaptation for Cross-Domain Person Re-IdentificationECCV 2020   
19Extending and Analyzing Self-Supervised Learning Across Domains (paper)ECCV 2020   
18Dual Mixup Regularized Learning for Adversarial Domain Adaptation (paper)ECCV 2020   
17Label Propagation with Augmented Anchors: A Simple Semi-Supervised Learning baseline for Unsupervised Domain Adaptation (paperECCV 2020codeSSL reguralization, Anchorsnew methods, good writings
16Do Adversarially Robust ImageNet Models Transfer Better?arvix 2020codeMany experiments 
15Visualizing Transfer Learningarvix 2020 interesting 
14A SURVEY ON DOMAIN ADAPTATION THEORY:LEARNING BOUNDS AND THEORETICAL GUARANTEES (paper)arvix 2020 theory 
13SpotTune: Transfer Learning through Adaptive Fine-tuning (paper)CVPR 2019code dynamic routing is a general method
12Parameter Transfer Unit for Deep Neural Networks (paper)PAKDD 2019 best paper  good idea, recommened to read
11Heterogeneous Domain Adaptation via Soft Transfer Network (paper)ACM MM 2019   
10Information-Theoretical Learning of Discriminative Clusters for Unsupervised Domain Adaptation (paper)ICML 2012   
9Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification (paper)arvix 2020  Good ideas
8Towards Recognizing Unseen Categories in Unseen Domains (paper)arvix 2020  new problems
7MiCo: Mixup Co-Training for Semi-Supervised Domain Adaptation (paper)arvix 2020  good framework
6Dynamic Knowledge Distillation for Black-box Hypothesis Transfer Learning (paperarvix 2020   
5Simultaneous Semantic Alignment Network for Heterogeneous Domain Adaptation (paper)ACM MM 2020code  
4Learning from a Complementary-label Source Domain: Theory and Algorithms(paper)arvix 2020code novel idea
3Class-Incremental Domain Adaptation(paper)ECCV 2020  new problems
2Class-incremental Learning via Deep Model Consolidation (paper)WACV 2020   
1Adversarial Graph Representation Adaptation for Cross-Domain Facial Expression Recognition (paper)ACM MM 2020  similar idea with us
0A Review of Single-Source Deep Unsupervised Visual Domain Adaptation paperarvix 2020 Reviewa good review! It contains many results of the state-of-the-art method

2) novel_papers on related fileds

numberTitleConference/journel + yearCodeKeywordsBenenit for us
14Self-Loop Uncertainty: A Novel Pseudo-Label for Semi-Supervised Medical Image Segmentation (paper)MICCAI 2020 ssl, pseudo label 
13Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning (paper)NIPS 2020 semi-supervised, weight smaplesit can be used in our work
12Safe semi-supervised learning: a brief introduction (paper)  safe sslnew concept, maybe useful for negative transfer
11Safe Deep Semi-Supervised Learning for Unseen-Class Unlabeled Data (paper)ICML 2020codessl, unseen classopen set, maybe useful for negative transfer
10(RECORD: Resource Constrained Semi-Supervised Learning under Distribution Shifpaper)KDD 2020 online, distribution shiftmaybe useful for negative transfer
9Adversarial Examples Improve Image Recognition (paper)CVPR 2020 Adversarial examples, image recognition, batch normalizationSame idea can be explored in DA
8Collaborative Graph Convolutional Networks: Unsupervised Learning Meets Semi-Supervised LearningAAAI 2020 unsupervised learning, semi-supervised learning 
7Self-supervised Label Augmentation via Input TransformationsICML 2020codeself-supervisedideas can be used to many tasks
6Learning with Multiple Complementary Labels (paper)ICML 2020   
5Deep Divergence Learning (paper)ICML 2020 divergence 
4Confidence-Aware Learning for Deep Neural Networks (paper)ICML 2020codeconfidence 
3Continual Learning in Human Activity Recognition:an Empirical Analysis of Regularization (paper)ICML workshopcodeContinual learning bechmark 
2Automated Phrase Mining from Massive Text Corpora (paper)    
1Adversarially-Trained Deep Nets Transfer Better(paperarvix 2020  new findings
0Learning to Combine: Knowledge Aggregation for Multi-Source Domain Adaptationarvix (paper)  same ideas with us

 

 

 

更多 DA awesome​​​​​​​

【必看】https://github.com/jindongwang/transferlearning

https://github.com/zhaoxin94/awesome-domain-adaptation#distance-based-methods

https://github.com/barebell/DA

 

入门参考

本部分内容适合初学者，将一些本领域中的经典论文按照时间线进行分类、梳理，分为浅层域适应、深度域适应、对抗域适应和域适应领域四部分。

针对每一部分，列举了3-4篇经典论文，建议详读这些经典论文，泛读这些经典论文的后续论文，并对其中的部分算法进行实现。

预期学习时间为2-3个月, 详细计划安排见入门参考

围绕这些论文，曾有一个相应的讨论班，相关的日程和资料如下：

小结

  • 迁移学习理论 ppt, pdf
  • Online Transfer Learning pdf

Excellent Scholars

新论文追踪

科研方法论

Presentation

  • 龙明盛 CCDM 2020 视频 , ppt
  • VALSE Webinar 20-19期 迁移学习 (个人非常推荐, 对新手不友好，对进阶有帮助，质量很高!) 视频, 报告简介
  • 龙明盛_NJU2019 Transfer Learning Theories and Algorithms ppt
  • 龙明盛 Valse 2019 Transfer Learning_From Algorithms to Theories and Back 视频 ppt
  • 游凯超 智源论坛 2019 领域适配前沿研究--场景、方法与模型选择 视频,ppt
  • 王玫 2019 deep_domain_adaptation 视频, ppt
  • 吴恩达 NIPS 2016 Nuts and bolts of building AI applications using Deep Learning 视频(需科学上网),ppt
Phoenixtree_DongZhao
关注
  • 1
    点赞
  • 踩
  • 11
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
专栏目录
迁移学习资料汇总
dejahu的博客
03-19 470
迁移学习 迁移学习这块不是我理解得简单得微调，微调这块得内容也是很讲究得，你冻结哪些参数，不冻结哪些参数都要进行指定 迁移学习的时候需要根据你的实际问题对网络结构进行调整，简单的模型只用调整最后的输出就行，但是对于复杂一点的问题而言， 我们还需要根据问题修改中间的卷积等层，或者说我们只是取出来这几个层的权重作为初始化的操作，来帮助我们进行训练， 而不是从随机初始化的权重来进行训练。 其实为了方便，...
迁移学习相关最新SCI文献
06-14
本资料集合包含了一系列关于迁移学习的最新SCI（科学引文索引）和Nature出版的英文文献，旨在为研究者和从业者提供宝贵的参考资源。 其中，"A deeply supervised residual network for HEp-2 cell classification ...
迁移学习教程
07-29
迁移学习详细教程，比较新的资料，值得细细阅读的资料。
迁移学习资料_迁移资料
cusi77914的博客
06-23 257
迁移或移动数据是一项常见的任务。 无论是通过网络将数据复制到新的文件系统，还是将逻辑卷复制到相同的卷组或不同的卷组中，或者只是创建文件系统的备份。 移动或复制数据的原因可能是由于性能问题，或者是由于当前环境中没有足够的空间而导致的数据总体增长。 有多种工具可用于上述数据移动任务，例如migratepv，cplv，tar，cpio，cp或rsync。 对于jfs，可以使用splitcopy，对于...
迁移学习 Transfer Learning(可能是目前最全的迁移学习资料库？)-附件资源
03-05
迁移学习 Transfer Learning(可能是目前最全的迁移学习资料库？)-附件资源
14 篇论文为你呈现「迁移学习」研究全貌 | 论文集精选 #04
Paper weekly
11-02 4845
PaperWeekly 是一个 AI 学术分享社区。这里聚集了大批一线 AI 学者，他们用精炼妙语推荐各自发现的优质论文。点击本文底部的「阅读原文」即刻加入社区，创建属于你的论文集。 这里是第 4 期论文集精选。 迁移学习对于人类来说，就是掌握举一反三的学习能力。对于计算机而言，所谓迁移学习，就是从一个或多个源任务（source task）中抽取知识和经验，然后将其应用于一个有相关性的目标
awesometypescript是一份收集整理了学习TypeScript资料集合
08-10
从 "awesome-typescript-master" 这个文件名来看，可能是一个包含各种 TypeScript 相关资源的项目仓库，其中包括示例代码、教程链接、社区资源等，可以帮助初学者快速上手，并为有经验的开发者提供进阶学习材料。...
元学习研究综述.docx
06-02
但是，迁移学习侧重于在实例空间进行训练，通过优化单个整体任务作为转移源来训练基础模型，而元学习是在任务空间中进行训练，每当模型尝试学习某项任务时，无论成功与否，模型都会获得有用的经验，整合这些经验形成...
深度学习资料大全002
04-07
这个压缩包“深度学习资料大全002”显然是一份包含了丰富深度学习资源的集合，源自某著名大学，适合于对深度学习感兴趣的学者和学生进行深入学习和研究。 首先，我们来看一下提供的文件列表： 1. 200310_Backprop....
第六章学习迁移共3页.pdf.zip
最新发布
11-25
学习迁移，也称为迁移学习，在信息技术和教育领域中，是指一个人从一个学习情境中获得的知识、技能或策略可以应用到另一个不同的但相关的学习情境中。这种能力在计算机科学，尤其是机器学习中具有重要意义，因为模型...
迁移学习相关论文和个人总结
11-10
有关迁移学习的相关论文和资料，并且附加上个人的总结。
Graph-Optimal-Transport:ICML 2020代码“跨域对齐的图形最佳传输”
03-09
跨域对齐的图最优传输框架 概括 在这项工作中，应用了Gromov-Wasserstein和Wasserstein距离来提高不同的跨域对齐任务的性能，例如VQA，机器翻译等。 象征性 注意，VQA模型是在基础上开发的。 并在建立了机器翻译模型。 引文 如果您发现该代码有用，请引用我们的ICML 2020论文。 @article{chen2020graph, title={Graph Optimal Transport for Cross-Domain Alignment}, author={Chen, Liqun and Gan, Zhe and Cheng, Yu and Li, Linjie and Carin, Lawrence and Liu, Jingjing}, journal={ICML}, year={2020} }
Label Efficient Semi-Supervised Learning via Graph Filtering.pdf
08-09
Graph-based methods have been demonstrated as one of the most effective approaches for semi-supervised learning,astheycanexploittheconnectivitypatternsbetweenlabeled and unlabeled data samples to improve learning performance. However, existing graph-based methods either are limited in their ability to jointly model graph structures and data features, such as the classical label propagation methods, or require a considerable amount of labeled data for training and validation due to high model complexity, such as the recent neural-network-based methods. In this paper, we address label efficient semi-supervised learning from a graph filtering perspective. Specifically, we propose a graph filtering framework that injects graph similarity into data features by taking them as signals on the graph and applying a low-pass graph filter to extract useful data representationsforclassification,wherelabelefficiencycan be achieved by conveniently adjusting the strength of the graph filter. Interestingly, this framework unifies two seemingly very different methods – label propagation and graph convolutional networks. Revisiting them under the graph filtering framework leads to new insights that improve their modelingcapabilitiesandreducemodelcomplexity. Experiments on various semi-supervised classification tasks on four citation networks and one knowledge graph and one semi-supervised regression task for zero-shot image recognition validate our findings and proposals.
matlab2017a代码-ImageNet-Models-on-Domain-Adaptation:ImageNet领域适应模型
05-28
matlab2017a代码ImageNet模型选择对域适应的影响 该目录包含纸张代码，在2020年IEEE冬季计算机视觉应用程序研讨会（WACVW）中被接受。 以及纸上的代码。 如有任何疑问，请发送电子邮件至 参考 如果您觉得有帮助，请引用为： Zhang, Youshan, and Brian D. Davison. (2020). Impact of ImageNet Model Selection on Domain Adaptation. In 2020 IEEE Winter Applications of Computer Vision Workshops (WACVW). Zhang, Y., & Davison, BD (2019). Modified distribution alignment for domain adaptation with pre-trained inception resnet. arXiv preprint arXiv:1904.02322. 或以bibtex样式： @article{zhang2020impact, title={Imp
Pytorch官方教程——迁移学习
DeepRoad
03-11 604
这篇文章翻译自 Pytorch 官方教程 Transfer Learning Tutorial 原作者：Sasank Chilamkurthy Note: 点击这里下载完整示例代码 在这篇教程中，你将会学到如何利用迁移学习来训练你的网络。你可以通过 cs231n notes 了解更多关于迁移学习的信息。 引用 cs231n notes 中的一段话 在实践中，很少有人会从头开始训练一个卷积...
深度学习+迁移学习资料整理
XU_RAL_KO的博客
11-24 241
文章目录前言一、Python机器学习1.1 sklearn库的学习二、深度学习框架三、工具整理 前言 对在个人学习过程中收集到的资料进行整理，仅供参考，持续更新（收藏=学会）。 一、Python机器学习 1.1 sklearn库的学习 官方文档地址：官方文档跳转 使用python中的sklearn扩展库，可以利用其提供的机器学习算法实现特征子集的筛选 学习参考链接： 1.Python机器学习笔记：sklearn库的学习——战争热诚整理，很详细 二、深度学习框架 三、工具整理 ...
Coursera | Andrew Ng (03-week2-2.7)—迁移学习
Self Improvement Lab
01-27 1016
该系列仅在原课程基础上部分知识点添加个人学习笔记，或相关推导补充等。如有错误，还请批评指教。在学习了 Andrew Ng 课程的基础上，为了更方便的查阅复习，将其整理成文字。因本人一直在学习英语，所以该系列以英文为主，同时也建议读者以英文为主，中文辅助，以便后期进阶时，为学习相关领域的学术论文做铺垫。- ZJ Coursera 课程 |deeplearning.ai |网易云课堂
关于迁移学习很好的文章
yuanxue18的博客
03-11 337
https://zhuanlan.zhihu.com/p/47991650
深度强化学习：全面综述与资源集合
3. 迁移学习：智能体可以将已学技能应用到新任务，减少训练时间。 4. 多智能体RL：处理多个智能体相互作用的问题，如合作或竞争。 5. 层次强化学习：通过分层次的任务分解，使学习更具结构化和高效。 6. 学习学习：...

“相关推荐”对你有帮助么？

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值