Deep Transfer Learning with Joint Adaptation Networks

Published in

International Conference on Machine Learning(2016)

External Links:

Generate Graph

TL;DR

JAN is presented, which learn a transfer network by aligning the joint distributions of multiple domain-specific layers across domains based on a joint maximum mean discrepancy (JMMD) criterion.

Abstract

Deep networks have been successfully applied to learn transferable features for adapting models from a source domain to a different target domain. In this paper, we present joint adaptation networks (JAN), which learn a transfer network by aligning the joint distributions of multiple domain-specific layers across domains based on a joint maximum mean discrepancy (JMMD) criterion. Adversarial training strategy is adopted to maximize JMMD such that the distributions of the source and target domains are made more distinguishable. Learning can be performed by stochastic gradient descent with the gradients computed by back-propagation in linear-time. Experiments testify that our model yields state of the art results on standard datasets.

Authors

Mingsheng Long

8 papers

Jianmin Wang

10 papers

Michael I. Jordan

14 papers

References46 items

Wasserstein Generative Adversarial Networks

Adversarial Discriminative Domain Adaptation

Domain Separation Networks

Domain Adaptation with Conditional Transferable Components

Unsupervised Domain Adaptation with Residual Transfer Networks

Deep Transfer Learning with Joint Adaptation Networks

Published in

International Conference on Machine Learning(2016)

External Links:

Generate Graph

TL;DR

JAN is presented, which learn a transfer network by aligning the joint distributions of multiple domain-specific layers across domains based on a joint maximum mean discrepancy (JMMD) criterion.

Abstract

Authors

Mingsheng Long

8 papers

Jianmin Wang

10 papers

Michael I. Jordan

14 papers

References46 items

Wasserstein Generative Adversarial Networks

Adversarial Discriminative Domain Adaptation

Domain Separation Networks

Domain Adaptation with Conditional Transferable Components

Unsupervised Domain Adaptation with Residual Transfer Networks

Deep Residual Learning for Image Recognition

Simultaneous Deep Transfer Across Domains and Tasks

On the High Dimensional Power of a Linear-Time Two Sample Test under Mean-shift Alternatives

Learning Transferable Features with Deep Adaptation Networks

Deep Domain Confusion: Maximizing for Domain Invariance

Flexible Transfer Learning under Support and Model Shift

How transferable are features in deep neural networks?

Unsupervised Domain Adaptation by Backpropagation

Going deeper with convolutions

Very Deep Convolutional Networks for Large-Scale Image Recognition

LSDA: Large Scale Detection through Adaptation

Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks

Robust Low Rank Kernel Embeddings of Multivariate Distributions

DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition

Domain Adaptation under Target and Conditional Shift

Connecting the Dots with Landmarks: Discriminatively Learning Domain-Invariant Features for Unsupervised Domain Adaptation

Kernel Embeddings of Conditional Distributions: A Unified Kernel Framework for Nonparametric Inference in Graphical Models

ImageNet classification with deep convolutional neural networks

Equivalence of distance-based and RKHS-based statistics in hypothesis testing

Representation Learning: A Review and New Perspectives

Geodesic flow kernel for unsupervised domain adaptation

A Kernel Two-Sample Test

Domain Transfer Multiple Kernel Learning

Domain adaptation for object recognition: An unsupervised approach

Domain Adaptation for Large-Scale Sentiment Classification: A Deep Learning Approach

Natural Language Processing (Almost) from Scratch

A Survey on Transfer Learning

Cross Validation Framework to Choose amongst Models and Datasets for Transfer Learning

Adapting Visual Category Models to New Domains

Hilbert Space Embeddings of Hidden Markov Models

A theory of learning from different domains

Kernel Choice and Classifiability for RKHS Embeddings of Probability Distributions

Hilbert Space Embeddings and Metrics on Probability Measures

Domain Adaptation via Transfer Component Analysis

Hilbert space embeddings of conditional distributions with applications to dynamical systems

Dataset Shift in Machine Learning

Domain Adaptation: Learning Bounds and Algorithms

Direct Importance Estimation with Model Selection and Its Application to Covariate Shift Adaptation

A Hilbert Space Embedding for Distributions

Correcting Sample Selection Bias by Unlabeled Data

GENERATIVE ADVERSARIAL NETS

Field of Study

Computer ScienceMathematics

Journal Information

Name

ArXiv

Volume

abs/1605.06636

Venue Information

Name

International Conference on Machine Learning

Type

conference

URL

https://icml.cc/

Alternate Names

ICML
Int Conf Mach Learn