Deep Domain Confusion: Maximizing for Domain Invariance

Abstract

Recent reports suggest that a generic supervised deep CNN model trained on a large-scale dataset reduces, but does not remove, dataset bias on a standard benchmark. Fine-tuning deep models in a new domain can require a significant amount of data, which for many applications is simply not available. We propose a new CNN architecture which introduces an adaptation layer and an additional domain confusion loss, to learn a representation that is both semantically meaningful and domain invariant. We additionally show that a domain confusion metric can be used for model selection to determine the dimension of an adaptation layer and the best position for the layer in the CNN architecture. Our proposed adaptation method offers empirical performance which exceeds previously published results on a standard benchmark visual domain adaptation task.

References in corpus (1)

• Frustratingly Easy Domain Adaptation

Cited by in corpus (224)

• Learning Transferable Features with Deep Adaptation Networks
• Unsupervised Domain Adaptation by Backpropagation
• Deep Subdomain Adaptation Network for Image Classification
• CyCADA: Cycle-Consistent Adversarial Domain Adaptation
• Domain Separation Networks
• Adversarial Discriminative Domain Adaptation
• Aligning Domain-specific Distribution and Classifier for Cross-domain Classification from Multiple Sources
• Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation
• Maximum Density Divergence for Domain Adaptation
• Like What You Like: Knowledge Distill via Neuron Selectivity Transfer
• Domain Adaptation for Visual Applications: A Comprehensive Survey
• Multi-Representation Adaptation Network for Cross-domain Image Classification
• Bridging Theory and Algorithm for Domain Adaptation
• Weakly Supervised Adversarial Domain Adaptation for Semantic Segmentation in Urban Scenes
• Few-Shot Adversarial Domain Adaptation
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• Collaborative Unsupervised Domain Adaptation for Medical Image Diagnosis
• A Berkeley View of Systems Challenges for AI
• Deep Residual Correction Network for Partial Domain Adaptation
• Learning Disentangled Semantic Representation for Domain Adaptation
• Contrastive Adaptation Network for Unsupervised Domain Adaptation
• Cross-regional oil palm tree counting and detection via multi-level attention domain adaptation network
• Complementary Pseudo Labels For Unsupervised Domain Adaptation On Person Re-identification
• Generalized Domain Conditioned Adaptation Network
• Partial Transfer Learning with Selective Adversarial Networks
• Multi-modal Deep Analysis for Multimedia
• Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
• Domain Adaptation with Asymmetrically-Relaxed Distribution Alignment
• COVID-DA: Deep Domain Adaptation from Typical Pneumonia to COVID-19
• Label Efficient Learning of Transferable Representations across Domains and Tasks
• A Survey on Transfer Learning in Natural Language Processing
• Transferrable Prototypical Networks for Unsupervised Domain Adaptation
• PnP-AdaNet: Plug-and-Play Adversarial Domain Adaptation Network with a Benchmark at Cross-modality Cardiac Segmentation
• Object Detection Using Deep CNNs Trained on Synthetic Images
• Domain Adaptation Meets Zero-Shot Learning: An Annotation-Efficient Approach to Multi-Modality Medical Image Segmentation
• Learning Target Domain Specific Classifier for Partial Domain Adaptation
• Knowledge Adaptation: Teaching to Adapt
• Domain-Symmetric Networks for Adversarial Domain Adaptation
• Domain Adaptation for Semantic Segmentation with Maximum Squares Loss
• Learning to Transfer Examples for Partial Domain Adaptation
• Feature Alignment and Restoration for Domain Generalization and Adaptation
• AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive Person Re-identification
• Smartphone Sensor-based Human Activity Recognition Robust to Different Sampling Rates
• An Unsupervised Domain Adaptation Model based on Dual-module Adversarial Training
• Multi-source Domain Adaptation for Text-independent Forensic Speaker Recognition
• Discriminative Radial Domain Adaptation
• Scalable Transfer Learning with Expert Models
• Robust Local Preserving and Global Aligning Network for Adversarial Domain Adaptation
• Domain-Specific Batch Normalization for Unsupervised Domain Adaptation
• Image to Image Translation for Domain Adaptation
• A Prototype-Oriented Framework for Unsupervised Domain Adaptation
• Joint Contrastive Learning for Unsupervised Domain Adaptation
• Understanding Self-Training for Gradual Domain Adaptation
• Rethink Maximum Mean Discrepancy for Domain Adaptation
• Iterative self-transfer learning: A general methodology for response time-history prediction based on small dataset
• Source-free Domain Adaptation via Avatar Prototype Generation and Adaptation
• Correlation Alignment by Riemannian Metric for Domain Adaptation
• Domain Adaptive Relational Reasoning for 3D Multi-Organ Segmentation
• Prototypical Cross-domain Self-supervised Learning for Few-shot Unsupervised Domain Adaptation
• Developing Personalized Models of Blood Pressure Estimation from Wearable Sensors Data Using Minimally-trained Domain Adversarial Neural Networks
• Dual Mixup Regularized Learning for Adversarial Domain Adaptation
• Synthetic to Real Adaptation with Generative Correlation Alignment Networks
• Sensor-invariant Fingerprint ROI Segmentation Using Recurrent Adversarial Learning
• Modified Distribution Alignment for Domain Adaptation with Pre-trained Inception ResNet
• Entropy Minimization vs. Diversity Maximization for Domain Adaptation
• Progressive Domain Expansion Network for Single Domain Generalization
• A Multi-task Deep Network for Person Re-identification
• Conditional out-of-sample generation for unpaired data using trVAE
• Attention Regularized Laplace Graph for Domain Adaptation
• Discriminative and Geometry Aware Unsupervised Domain Adaptation
• EllSeg-Gen, towards Domain Generalization for head-mounted eyetracking
• ConDA: Continual Unsupervised Domain Adaptation
• Network Architecture Search for Domain Adaptation
• Dynamic Weighted Learning for Unsupervised Domain Adaptation
• Label Propagation with Augmented Anchors: A Simple Semi-Supervised Learning baseline for Unsupervised Domain Adaptation
• Contextual-Relation Consistent Domain Adaptation for Semantic Segmentation
• Domain Alignment with Triplets
• Unsupervised Domain Adaptation via Structured Prediction Based Selective Pseudo-Labeling
• Exploiting Domain Transferability for Collaborative Inter-level Domain Adaptive Object Detection
• Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation
• Learning to Combine: Knowledge Aggregation for Multi-Source Domain Adaptation
• Associative Partial Domain Adaptation
• Gradient Regularized Contrastive Learning for Continual Domain Adaptation
• DART: Domain-Adversarial Residual-Transfer Networks for Unsupervised Cross-Domain Image Classification
• VisDA-2021 Competition Universal Domain Adaptation to Improve Performance on Out-of-Distribution Data
• Dual Adversarial Domain Adaptation
• Few-shot Unsupervised Domain Adaptation with Image-to-class Sparse Similarity Encoding
• Adversarial Domain Adaptation with Domain Mixup
• Multi-Objective Meta Learning
• A Unified Joint Maximum Mean Discrepancy for Domain Adaptation
• Collaborative Training of Balanced Random Forests for Open Set Domain Adaptation
• Unsupervised domain adaptation in brain lesion segmentation with adversarial networks
• Learning Cross-domain Generalizable Features by Representation Disentanglement
• Fine-grained Recognition in the Wild: A Multi-Task Domain Adaptation Approach
• A Curriculum Domain Adaptation Approach to the Semantic Segmentation of Urban Scenes
• DAMIA: Leveraging Domain Adaptation as a Defense against Membership Inference Attacks
• Known-class Aware Self-ensemble for Open Set Domain Adaptation
• On Universal Black-Box Domain Adaptation
• AdapNet: Adaptability Decomposing Encoder-Decoder Network for Weakly Supervised Action Recognition and Localization
• HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation
• Adversarial-Learned Loss for Domain Adaptation
• Joint Feature Distribution Alignment Learning for NIR-VIS and VIS-VIS Face Recognition
• Beyond Fine Tuning: A Modular Approach to Learning on Small Data
• Deep causal representation learning for unsupervised domain adaptation
• Data Augmentation with norm-VAE for Unsupervised Domain Adaptation
• Modeling the Sequential Dependence among Audience Multi-step Conversions with Multi-task Learning in Targeted Display Advertising
• Counterfactual Invariance to Spurious Correlations: Why and How to Pass Stress Tests
• Learning to Match Distributions for Domain Adaptation
• MetaAlign: Coordinating Domain Alignment and Classification for Unsupervised Domain Adaptation
• Domain Generalization via Semi-supervised Meta Learning
• Between-Domain Instance Transition Via the Process of Gibbs Sampling in RBM
• Learning Disentangled Representations for Time Series
• Unsupervised Visual Domain Adaptation: A Deep Max-Margin Gaussian Process Approach
• Generalized Zero and Few-Shot Transfer for Facial Forgery Detection
• Privacy-preserving Transfer Learning via Secure Maximum Mean Discrepancy
• Bi-Classifier Determinacy Maximization for Unsupervised Domain Adaptation
• Cross-Entropy Adversarial View Adaptation for Person Re-identification
• Discriminative Adversarial Domain Adaptation
• Deep Domain Adaptation by Geodesic Distance Minimization
• Weakly Supervised Open-set Domain Adaptation by Dual-domain Collaboration
• Unsupervised Domain Adaptation for Image Classification via Structure-Conditioned Adversarial Learning
• I3Net: Implicit Instance-Invariant Network for Adapting One-Stage Object Detectors
• An Extended Framework for Marginalized Domain Adaptation
• Single Domain Dynamic Generalization for Iris Presentation Attack Detection
• DRANet: Disentangling Representation and Adaptation Networks for Unsupervised Cross-Domain Adaptation
• Joint Semantic Domain Alignment and Target Classifier Learning for Unsupervised Domain Adaptation
• Moment-Based Domain Adaptation: Learning Bounds and Algorithms
• Unsupervised Cross-domain Image Classification by Distance Metric Guided Feature Alignment
• Deeply Aligned Adaptation for Cross-domain Object Detection
• Federated Multi-Target Domain Adaptation
• JOKR: Joint Keypoint Representation for Unsupervised Cross-Domain Motion Retargeting
• Energy Confused Adversarial Metric Learning for Zero-Shot Image Retrieval and Clustering
• Locality Preserving Joint Transfer for Domain Adaptation
• Multi-Source Domain Adaptation for Text Classification via DistanceNet-Bandits
• Unsupervised Domain Adaptation with Temporal-Consistent Self-Training for 3D Hand-Object Joint Reconstruction
• AutoAdapt: Automated Segmentation Network Search for Unsupervised Domain Adaptation
• Mitigating Uncertainty of Classifier for Unsupervised Domain Adaptation
• Self-adaptive Re-weighted Adversarial Domain Adaptation
• Learning a Domain Classifier Bank for Unsupervised Adaptive Object Detection
• Domain Conditioned Adaptation Network
• CODE-AE: A Coherent De-confounding Autoencoder for Predicting Patient-Specific Drug Response From Cell Line Transcriptomics
• Uncertainty-Aware Model Adaptation for Unsupervised Cross-Domain Object Detection
• Robustified Domain Adaptation
• Towards Self-similarity Consistency and Feature Discrimination for Unsupervised Domain Adaptation
• Mitigating the Impact of Speech Recognition Errors on Chatbot using Sequence-to-Sequence Model
• Light-weight Calibrator: a Separable Component for Unsupervised Domain Adaptation
• Learning Domain Adaptive Features with Unlabeled Domain Bridges
• Improving Unsupervised Domain Adaptation by Reducing Bi-level Feature Redundancy
• Learning Invariant Representation with Consistency and Diversity for Semi-supervised Source Hypothesis Transfer
• SODA: Detecting Covid-19 in Chest X-rays with Semi-supervised Open Set Domain Adaptation
• Air-Writing Translater: A Novel Unsupervised Domain Adaptation Method for Inertia-Trajectory Translation of In-air Handwriting
• Low-Resource Sequence Labeling via Unsupervised Multilingual Contextualized Representations
• Conditional Coupled Generative Adversarial Networks for Zero-Shot Domain Adaptation
• Unsupervised Domain Adaptation using Deep Networks with Cross-Grafted Stacks
• Exploring Category-Agnostic Clusters for Open-Set Domain Adaptation
• Disentanglement Then Reconstruction: Learning Compact Features for Unsupervised Domain Adaptation
• StereoGAN: Bridging Synthetic-to-Real Domain Gap by Joint Optimization of Domain Translation and Stereo Matching
• Sequential Learning for Domain Generalization
• Discriminative Clustering for Robust Unsupervised Domain Adaptation
• Image Alignment in Unseen Domains via Domain Deep Generalization
• Learning Invariant Representations across Domains and Tasks
• Domain Impression: A Source Data Free Domain Adaptation Method
• Unsupervised Domain Adaptive Learning via Synthetic Data for Person Re-identification
• Multi-source Few-shot Domain Adaptation
• Transporting Causal Mechanisms for Unsupervised Domain Adaptation
• Compositional Models: Multi-Task Learning and Knowledge Transfer with Modular Networks
• Joint Semi-supervised 3D Super-Resolution and Segmentation with Mixed Adversarial Gaussian Domain Adaptation
• Learning from Synthetic Data for Opinion-free Blind Image Quality Assessment in the Wild
• Stable Distribution Alignment Using the Dual of the Adversarial Distance
• Cross-domain Imitation from Observations
• OmniPrint: A Configurable Printed Character Synthesizer
• Impact of ImageNet Model Selection on Domain Adaptation
• Domain2Vec: Domain Embedding for Unsupervised Domain Adaptation
• Unsupervised Open Domain Recognition by Semantic Discrepancy Minimization
• On the Robustness of Domain Adaption to Adversarial Attacks
• Exposing Semantic Segmentation Failures via Maximum Discrepancy Competition
• Self-Supervised Dynamic Networks for Covariate Shift Robustness
• Adversarial Learning for Zero-shot Domain Adaptation
• Range Adaptation for 3D Object Detection in LiDAR
• Enlarging Discriminative Power by Adding an Extra Class in Unsupervised Domain Adaptation
• Learning Condensed and Aligned Features for Unsupervised Domain Adaptation Using Label Propagation
• Against Adversarial Learning: Naturally Distinguish Known and Unknown in Open Set Domain Adaptation
• Divergence Optimization for Noisy Universal Domain Adaptation
• Unsupervised Domain Expansion for Visual Categorization
• Discriminative Noise Robust Sparse Orthogonal Label Regression-based Domain Adaptation
• Fighting Copycat Agents in Behavioral Cloning from Observation Histories
• Open-Set Crowdsourcing using Multiple-Source Transfer Learning
• Unsupervised Domain Adaptation Network with Category-Centric Prototype Aligner for Biomedical Image Segmentation
• Multi-sequence Cardiac MR Segmentation with Adversarial Domain Adaptation Network
• Appearance invariance in convolutional networks with neighborhood similarity
• Domain Consistency Regularization for Unsupervised Multi-source Domain Adaptive Classification
• Unsupervised Domain Adaptation with Variational Approximation for Cardiac Segmentation
• Deep Doubly Supervised Transfer Network for Diagnosis of Breast Cancer with Imbalanced Ultrasound Imaging Modalities
• Learning to see across Domains and Modalities
• Syn2Real: Forgery Classification via Unsupervised Domain Adaptation
• Efficient Transfer Learning via Joint Adaptation of Network Architecture and Weight
• DWMD: Dimensional Weighted Orderwise Moment Discrepancy for Domain-specific Hidden Representation Matching
• Learning Task-oriented Disentangled Representations for Unsupervised Domain Adaptation
• Multi-component Image Translation for Deep Domain Generalization
• Transducer Adaptive Ultrasound Volume Reconstruction
• Universal Multi-Source Domain Adaptation
• Mixed Set Domain Adaptation
• One Network for Multi-Domains: Domain Adaptive Hashing with Intersectant Generative Adversarial Network
• All-Weather Object Recognition Using Radar and Infrared Sensing
• Deep Least Squares Alignment for Unsupervised Domain Adaptation
• Domain Adaptation for Rare Classes Augmented with Synthetic Samples
• Unsupervised Domain Adaptive Object Detection using Forward-Backward Cyclic Adaptation
• Agile Domain Adaptation
• Evaluation of Deep Neural Network Domain Adaptation Techniques for Image Recognition
• Parameter Reference Loss for Unsupervised Domain Adaptation
• Task Guided Compositional Representation Learning for ZDA
• Domain Adaptation by Maximizing Population Correlation with Neural Architecture Search
• On-target Adaptation
• Contrastive Domain Adaptation for Question Answering using Limited Text Corpora
• Domain Adaptation for Real-World Single View 3D Reconstruction
• Meta Self-Learning for Multi-Source Domain Adaptation: A Benchmark
• Zero-Round Active Learning
• Ensemble Multi-Source Domain Adaptation with Pseudolabels
• Wasserstein Distance Guided Cross-Domain Learning
• Exploring Dropout Discriminator for Domain Adaptation
• From Real to Synthetic and Back: Synthesizing Training Data for Multi-Person Scene Understanding
• Cross-domain error minimization for unsupervised domain adaptation
• Enhanced Separable Disentanglement for Unsupervised Domain Adaptation