VisDA: The Visual Domain Adaptation Challenge

Abstract

We present the 2017 Visual Domain Adaptation (VisDA) dataset and challenge, a large-scale testbed for unsupervised domain adaptation across visual domains. Unsupervised domain adaptation aims to solve the real-world problem of domain shift, where machine learning models trained on one domain must be transferred and adapted to a novel visual domain without additional supervision. The VisDA2017 challenge is focused on the simulation-to-reality shift and has two associated tasks: image classification and image segmentation. The goal in both tracks is to first train a model on simulated, synthetic data in the source domain and then adapt it to perform well on real image data in the unlabeled test domain. Our dataset is the largest one to date for cross-domain object classification, with over 280K images across 12 categories in the combined training, validation and testing domains. The image segmentation dataset is also large-scale with over 30K images across 18 categories in the three domains. We compare VisDA to existing cross-domain adaptation datasets and provide a baseline performance analysis using various domain adaptation models that are currently popular in the field.

References in corpus (6)

• FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
• Equivalence of distance-based and RKHS-based statistics in hypothesis testing
• Texture Networks: Feed-forward Synthesis of Textures and Stylized Images
• Domain Adaptation for Visual Applications: A Comprehensive Survey
• A Large Dataset of Object Scans
• A Testbed for Cross-Dataset Analysis

Cited by in corpus (149)

• Deep Subdomain Adaptation Network for Image Classification
• Domain Generalization: A Survey
• DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptation
• Bridging Theory and Algorithm for Domain Adaptation
• Universal Domain Adaptation through Self Supervision
• Maximum Classifier Discrepancy for Unsupervised Domain Adaptation
• Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation
• CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation
• Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation
• Unsupervised Multi-Class Domain Adaptation: Theory, Algorithms, and Practice
• Contrastive Adaptation Network for Unsupervised Domain Adaptation
• Unsupervised Domain Adaptation via Discriminative Manifold Propagation
• Domain Adaptation with Conditional Distribution Matching and Generalized Label Shift
• Cycle Self-Training for Domain Adaptation
• Learning Transferable Parameters for Unsupervised Domain Adaptation
• Transferrable Prototypical Networks for Unsupervised Domain Adaptation
• Moment Matching for Multi-Source Domain Adaptation
• Larger Norm More Transferable: An Adaptive Feature Norm Approach for Unsupervised Domain Adaptation
• Uncertainty-Induced Transferability Representation for Source-Free Unsupervised Domain Adaptation
• Adversarial Domain Adaptation with Prototype-Based Normalized Output Conditioner
• Test-Time Adaptation to Distribution Shift by Confidence Maximization and Input Transformation
• Unsupervised Domain Adaptation of Black-Box Source Models
• Robust Optimal Transport with Applications in Generative Modeling and Domain Adaptation
• Rethinking Distributional Matching Based Domain Adaptation
• Unbalanced minibatch Optimal Transport; applications to Domain Adaptation
• Cross-domain Self-supervised Learning for Domain Adaptation with Few Source Labels
• The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models
• Drop to Adapt: Learning Discriminative Features for Unsupervised Domain Adaptation
• An Unsupervised Domain Adaptation Model based on Dual-module Adversarial Training
• Source Free Domain Adaptation with Image Translation
• Transfer Learning with Label Noise
• Minimum Class Confusion for Versatile Domain Adaptation
• Domain-Specific Batch Normalization for Unsupervised Domain Adaptation
• Uncertainty-Aware Pseudo-Label Filtering for Source-Free Unsupervised Domain Adaptation
• Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer
• Domain Adaptation without Source Data
• TVT: Transferable Vision Transformer for Unsupervised Domain Adaptation
• A Review of Single-Source Deep Unsupervised Visual Domain Adaptation
• Joint Contrastive Learning for Unsupervised Domain Adaptation
• RAIN: RegulArization on Input and Network for Black-Box Domain Adaptation
• Transformer-Based Source-Free Domain Adaptation
• Domain Randomization and Pyramid Consistency: Simulation-to-Real Generalization without Accessing Target Domain Data
• Lessons from a Space Lab -- An Image Acquisition Perspective
• Transferable Calibration with Lower Bias and Variance in Domain Adaptation
• Contrastive Syn-to-Real Generalization
• Unsupervised Domain Adaption Harnessing Vision-Language Pre-training
• Towards Inheritable Models for Open-Set Domain Adaptation
• MixNorm: Test-Time Adaptation Through Online Normalization Estimation
• Characterizing and Avoiding Negative Transfer
• Prototypical Partial Optimal Transport for Universal Domain Adaptation
• Prototypical Cross-domain Self-supervised Learning for Few-shot Unsupervised Domain Adaptation
• Source-free Domain Adaptation via Avatar Prototype Generation and Adaptation
• Semi-supervised Models are Strong Unsupervised Domain Adaptation Learners
• MixStyle Neural Networks for Domain Generalization and Adaptation
• Gradually Vanishing Bridge for Adversarial Domain Adaptation
• OVANet: One-vs-All Network for Universal Domain Adaptation
• Cross-domain Contrastive Learning for Unsupervised Domain Adaptation
• Open-Set Hypothesis Transfer with Semantic Consistency
• Simulations of Common Unsupervised Domain Adaptation Algorithms for Image Classification
• ConDA: Continual Unsupervised Domain Adaptation
• Dynamic Weighted Learning for Unsupervised Domain Adaptation
• Label Propagation with Augmented Anchors: A Simple Semi-Supervised Learning baseline for Unsupervised Domain Adaptation
• Learning Invariant Representations and Risks for Semi-supervised Domain Adaptation
• Angular Visual Hardness
• Associative Partial Domain Adaptation
• MiniMax Entropy Network: Learning Category-Invariant Features for Domain Adaptation
• Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation
• Generation, augmentation, and alignment: A pseudo-source domain based method for source-free domain adaptation
• Transfering Low-Frequency Features for Domain Adaptation
• FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation
• Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptation
• Domain Adaptation with Auxiliary Target Domain-Oriented Classifier
• Adversarial Domain Adaptation with Domain Mixup
• Open Domain Generalization with Domain-Augmented Meta-Learning
• Effective Label Propagation for Discriminative Semi-Supervised Domain Adaptation
• A Curriculum Domain Adaptation Approach to the Semantic Segmentation of Urban Scenes
• Exploiting Both Domain-specific and Invariant Knowledge via a Win-win Transformer for Unsupervised Domain Adaptation
• When Invariant Representation Learning Meets Label Shift: Insufficiency and Theoretical Insights
• Nearest Neighborhood-Based Deep Clustering for Source Data-absent Unsupervised Domain Adaptation
• Sim-Real Joint Reinforcement Transfer for 3D Indoor Navigation
• Class-imbalanced Domain Adaptation: An Empirical Odyssey
• DINE: Domain Adaptation from Single and Multiple Black-box Predictors
• A Sample Selection Approach for Universal Domain Adaptation
• Adversarial-Learned Loss for Domain Adaptation
• ToAlign: Task-oriented Alignment for Unsupervised Domain Adaptation
• Generalized Zero and Few-Shot Transfer for Facial Forgery Detection
• Unsupervised Visual Domain Adaptation: A Deep Max-Margin Gaussian Process Approach
• Learning to Match Distributions for Domain Adaptation
• Learning Smooth Representation for Unsupervised Domain Adaptation
• Normalized Wasserstein Distance for Mixture Distributions with Applications in Adversarial Learning and Domain Adaptation
• Contrastive Reasoning in Neural Networks
• Reducing the Covariate Shift by Mirror Samples in Cross Domain Alignment
• Unsupervised Domain Adaptation by Uncertain Feature Alignment
• Bi-Classifier Determinacy Maximization for Unsupervised Domain Adaptation
• Unsupervised Domain Adaptation for Image Classification via Structure-Conditioned Adversarial Learning
• BiPC: Bidirectional Probability Calibration for Unsupervised Domain Adaption
• Cycle-consistent Conditional Adversarial Transfer Networks
• Prototypical Cross-domain Knowledge Transfer for Cervical Dysplasia Visual Inspection
• Zero-shot Domain Adaptation Based on Attribute Information
• Large-Scale Historical Watermark Recognition: dataset and a new consistency-based approach
• Joint Semantic Domain Alignment and Target Classifier Learning for Unsupervised Domain Adaptation
• ECACL: A Holistic Framework for Semi-Supervised Domain Adaptation
• Improving Mini-batch Optimal Transport via Partial Transportation
• Locality Preserving Joint Transfer for Domain Adaptation
• A General Upper Bound for Unsupervised Domain Adaptation
• Improving Unsupervised Domain Adaptation by Reducing Bi-level Feature Redundancy
• Robustified Domain Adaptation
• Gradual Domain Adaptation via Self-Training of Auxiliary Models
• Multiple Classifiers Based Maximum Classifier Discrepancy for Unsupervised Domain Adaptation
• Visualizing Adapted Knowledge in Domain Transfer
• Contradictory Structure Learning for Semi-supervised Domain Adaptation
• Using Latent Codes for Class Imbalance Problem in Unsupervised Domain Adaptation
• Unsupervised Domain Adaptation using Generative Models and Self-ensembling
• Unsupervised domain adaptation via double classifiers based on high confidence pseudo label
• Unsupervised Domain Adaptation across FMCW Radar Configurations Using Margin Disparity Discrepancy
• Energy-Based Pseudo-Label Refining for Source-free Domain Adaptation
• What Does Rotation Prediction Tell Us about Classifier Accuracy under Varying Testing Environments?
• Mining Label Distribution Drift in Unsupervised Domain Adaptation
• Transporting Causal Mechanisms for Unsupervised Domain Adaptation
• Target Consistency for Domain Adaptation: when Robustness meets Transferability
• Adversarial Robustness for Unsupervised Domain Adaptation
• Stochastic Adversarial Gradient Embedding for Active Domain Adaptation
• Learning Invariant Representations across Domains and Tasks
• Disentanglement Then Reconstruction: Learning Compact Features for Unsupervised Domain Adaptation
• Domain2Vec: Domain Embedding for Unsupervised Domain Adaptation
• Divergence Optimization for Noisy Universal Domain Adaptation
• Removing Undesirable Feature Contributions Using Out-of-Distribution Data
• Towards Category and Domain Alignment: Category-Invariant Feature Enhancement for Adversarial Domain Adaptation
• Metric-Learning-Assisted Domain Adaptation
• Automatic Face Understanding: Recognizing Families in Photos
• Domain Adaptation by Maximizing Population Correlation with Neural Architecture Search
• Deep Least Squares Alignment for Unsupervised Domain Adaptation
• Certainty Volume Prediction for Unsupervised Domain Adaptation
• Deep Spherical Manifold Gaussian Kernel for Unsupervised Domain Adaptation
• Open Set Domain Adaptation for Image and Action Recognition
• Contrastive Vicinal Space for Unsupervised Domain Adaptation
• Dynamic Transfer for Multi-Source Domain Adaptation
• Frustratingly Easy Uncertainty Estimation for Distribution Shift
• Synthetic Data Are as Good as the Real for Association Knowledge Learning in Multi-object Tracking
• Transferable Semantic Augmentation for Domain Adaptation
• Enhanced Separable Disentanglement for Unsupervised Domain Adaptation
• Unsupervised Domain Adaptation via Discriminative Manifold Embedding and Alignment
• Leveraging Domain Adaptation for Low-Resource Geospatial Machine Learning
• Image-to-Image Translation of Synthetic Samples for Rare Classes
• Dynamic Domain Adaptation for Efficient Inference
• ADeLA: Automatic Dense Labeling with Attention for Viewpoint Adaptation in Semantic Segmentation
• Zero-Round Active Learning
• On-target Adaptation
• What Can Style Transfer and Paintings Do For Model Robustness?