ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation

2018/11/30 by Tuan-Hung Vu, Himalaya Jain, Vu, Tuan-Hung +7 · 145 citations
Computer Science · Medicine · #Advanced Neural Network Applications #COVID-19 diagnosis using AI #Domain Adaptation and Few-Shot Learning #cs.CV

paper · pdf · doi:10.48550/arxiv.1811.12833

Accepted in CVPR'19. Code is available at https://github.com/valeoai/ADVENT

arxiv created 2019/04/17 · arxiv updated 2019/04/19

Abstract

Semantic segmentation is a key problem for many computer vision tasks. While approaches based on convolutional neural networks constantly break new records on different benchmarks, generalizing well to diverse testing environments remains a major challenge. In numerous real world applications, there is indeed a large gap between data distributions in train and test domains, which results in severe performance loss at run-time. In this work, we address the task of unsupervised domain adaptation in semantic segmentation with losses based on the entropy of the pixel-wise predictions. To this end, we propose two novel, complementary methods using (i) entropy loss and (ii) adversarial loss respectively. We demonstrate state-of-the-art performance in semantic segmentation on two challenging "synthetic-2-real" set-ups and show that the approach can also be used for detection.

Cited by

• LabOR: Labeling Only if Required for Domain Adaptive Semantic Segmentation
• Domain Adaptation for Semantic Segmentation via Patch-Wise Contrastive Learning
• Unsupervised Domain Adaptation with Temporal-Consistent Self-Training for 3D Hand-Object Joint Reconstruction
• Source-Free Domain Adaptation for Semantic Segmentation
• Semantic Distribution-aware Contrastive Adaptation for Semantic Segmentation
• Spectral Unsupervised Domain Adaptation for Visual Recognition
• Collaborative Training between Region Proposal Localization and Classification for Domain Adaptive Object Detection
• OMUDA: Omni-level Masking for Unsupervised Domain Adaptation in Semantic Segmentation
• Semi-supervised Domain Adaptation for Semantic Segmentation
• DANNet: A One-Stage Domain Adaptation Network for Unsupervised Nighttime Semantic Segmentation
• Towards Robust Pseudo-Label Learning in Semantic Segmentation: An Encoding Perspective
• Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer
• Classes Matter: A Fine-grained Adversarial Approach to Cross-domain Semantic Segmentation
• WEDGE: Web-Image Assisted Domain Generalization for Semantic Segmentation
• Hybrid-Domain Adaptative Representation Learning for Gaze Estimation
• Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentation
• Mutual-GAN: Towards Unsupervised Cross-Weather Adaptation with Mutual Information Constraint
• Learning Texture Invariant Representation for Domain Adaptation of Semantic Segmentation
• Boosting Semi-supervised Image Segmentation with Global and Local Mutual Information Regularization
• TCSA-UDA: Text-Driven Cross-Semantic Alignment for Unsupervised Domain Adaptation in Medical Image Segmentation
• Decoupled Entropy Minimization
• Domain Adaptation for Semantic Segmentation with Maximum Squares Loss
• What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation
• FDA: Fourier Domain Adaptation for Semantic Segmentation
• Instance Adaptive Self-Training for Unsupervised Domain Adaptation
• Are We Hungry for 3D LiDAR Data for Semantic Segmentation? A Survey and Experimental Study
• Towards Stable Source-Free Domain Adaptive Semantic Segmentation
• Pixel-Level Cycle Association: A New Perspective for Domain Adaptive Semantic Segmentation
• SRPL-SFDA: SAM-Guided Reliable Pseudo-Labels for Source-Free Domain Adaptation in Medical Image Segmentation
• Fast Batch Nuclear-norm Maximization and Minimization for Robust Domain Adaptation
• UP2D: Uncertainty-aware Progressive Pseudo-label Denoising for Source-Free Domain Adaptive Medical Image Segmentation
• DPGLA: Bridging the Gap between Synthetic and Real Data for Unsupervised Domain Adaptation in 3D LiDAR Semantic Segmentation
• Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation
• Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation
• Tackling Catastrophic Forgetting and Background Shift in Continual Semantic Segmentation
• Source-Free Cross-Domain Continual Learning
• MATCH: Multi-faceted Adaptive Topo-Consistency for Semi-Supervised Histopathology Segmentation
• Source-Free Domain Adaptive Fundus Image Segmentation with Denoised Pseudo-Labeling
• Confidence Regularized Self-Training
• Bézier Meets Diffusion: Robust Generation Across Domains for Medical Image Segmentation
• Model Adaptation: Historical Contrastive Learning for Unsupervised Domain Adaptation without Source Data
• Anatomical feature‐guided semi‐supervised recognition of the vertebrobasilar artery for microvascular decompression
• Weakly-Supervised Cross-Domain Adaptation for Endoscopic Lesions Segmentation
• Source-Free Domain Adaptive Semantic Segmentation of Remote Sensing Images with Diffusion-Guided Label Enrichment
• Semi-MoE: Mixture-of-Experts meets Semi-Supervised Histopathology Segmentation
• FSDR: Frequency Space Domain Randomization for Domain Generalization
• Automatic Validation of Textual Attribute Values in E-commerce Catalog by Learning with Limited Labeled Data
• Adapt Everywhere: Unsupervised Adaptation of Point-Clouds and Entropy Minimisation for Multi-modal Cardiac Image Segmentation
• Addressing Failure Prediction by Learning Model Confidence
• Sparse-to-dense Feature Matching: Intra and Inter domain Cross-modal Learning in Domain Adaptation for 3D Semantic Segmentation
• Grad-CL: Source Free Domain Adaptation with Gradient Guided Feature Disalignment
• Uncertainty-Aware Unsupervised Domain Adaptation in Object Detection
• Integrating Information Theory and Adversarial Learning for Cross-modal Retrieval
• Generalizable Model-agnostic Semantic Segmentation via Target-specific Normalization
• Consistency Regularization with High-dimensional Non-adversarial Source-guided Perturbation for Unsupervised Domain Adaptation in Segmentation
• Generative AI for Enhanced Wildfire Detection: Bridging the Synthetic-Real Domain Gap
• Domain Adaptation via Feature Refinement
• Both Style and Fog Matter: Cumulative Domain Adaptation for Semantic Foggy Scene Understanding
• Efficient Semi-Supervised Gross Target Volume of Nasopharyngeal Carcinoma Segmentation via Uncertainty Rectified Pyramid Consistency
• Transferring and Regularizing Prediction for Semantic Segmentation
• One-shot Unsupervised Domain Adaptation with Personalized Diffusion Models
• Test-Time Adaptation to Distribution Shift by Confidence Maximization and Input Transformation
• Towards Fewer Annotations: Active Learning via Region Impurity and Prediction Uncertainty for Domain Adaptive Semantic Segmentation
• Cross-domain Object Detection through Coarse-to-Fine Feature Adaptation
• Unsupervised Domain Adaptation for Semantic Segmentation via Low-level Edge Information Transfer
• DatasetGAN: Efficient Labeled Data Factory with Minimal Human Effort
• Unified and Semantically Grounded Domain Adaptation for Medical Image Segmentation
• Hierarchical Adaptive networks with Task vectors for Test-Time Adaptation
• More Separable and Easier to Segment: A Cluster Alignment Method for Cross-Domain Semantic Segmentation
• Continual Unsupervised Domain Adaptation for Semantic Segmentation
• Rein++: Efficient Generalization and Adaptation for Semantic Segmentation with Vision Foundation Models
• Proto-EVFL: Enhanced Vertical Federated Learning via Dual Prototype with Extremely Unaligned Data
• Learning from Heterogeneous Structural MRI via Collaborative Domain Adaptation for Late-Life Depression Assessment
• Context-aware virtual adversarial training for anatomically-plausible segmentation
• Ensemble of Discriminators for Domain Adaptation in Multiple Sound Source 2D Localization
• AFRDA: Attentive Feature Refinement for Domain Adaptive Semantic Segmentation
• RarePlanes: Synthetic Data Takes Flight
• UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentation
• Learning from Extrinsic and Intrinsic Supervisions for Domain Generalization
• DAVOS: Semi-Supervised Video Object Segmentation via Adversarial Domain Adaptation
• Domain Adaptive Semantic Segmentation without Source Data
• Diversified Multi-prototype Representation for Semi-supervised Segmentation
• FARCLUSS: Fuzzy Adaptive Rebalancing and Contrastive Uncertainty Learning for Semi-Supervised Semantic Segmentation
• ESL: Entropy-guided Self-supervised Learning for Domain Adaptation in Semantic Segmentation
• Cross-Region Domain Adaptation for Class-level Alignment
• Learning Cross-modal Contrastive Features for Video Domain Adaptation
• Channel-wise Alignment for Adaptive Object Detection
• xMUDA: Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentation
• Bi-Classifier Determinacy Maximization for Unsupervised Domain Adaptation
• Transfer beyond the Field of View: Dense Panoramic Semantic Segmentation via Unsupervised Domain Adaptation
• Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-Supervision
• An Adversarial Perturbation Oriented Domain Adaptation Approach for Semantic Segmentation
• φ-Adapt: A Physics-Informed Adaptation Learning Approach to 2D Quantum Material Discovery
• Synergistic saliency and depth prediction for RGB-D saliency detection
• SG-LDM: Semantic-Guided LiDAR Generation via Latent-Aligned Diffusion
• Data-Efficient Framework for Real-world Multiple Sound Source 2D Localization
• AdvMIM: Adversarial Masked Image Modeling for Semi-Supervised Medical Image Segmentation
• Bi-Mix: Bidirectional Mixing for Domain Adaptive Nighttime Semantic Segmentation
• Semi-Supervised Multi-Modal Medical Image Segmentation for Complex Situations
• Phase Consistent Ecological Domain Adaptation
• Prmpt2Adpt: Prompt-Based Zero-Shot Domain Adaptation for Resource-Constrained Environments
• RealDriveSim: A Realistic Multi-Modal Multi-Task Synthetic Dataset for Autonomous Driving
• BIM Hyperreality: Data Synthesis Using BIM and Hyperrealistic Rendering for Deep Learning
• A Prototype-Oriented Framework for Unsupervised Domain Adaptation
• Exploiting Negative Learning for Implicit Pseudo Label Rectification in Source-Free Domain Adaptive Semantic Segmentation
• Domain Adaptation for Big Data in Agricultural Image Analysis: A Comprehensive Review
• Reliably Detecting Model Failures in Deployment Without Labels
• Adversarial Style Mining for One-Shot Unsupervised Domain Adaptation
• Coarse-to-Fine Domain Adaptive Semantic Segmentation with Photometric Alignment and Category-Center Regularization
• Style Normalization and Restitution for Domain Generalization and Adaptation
• Cross-denoising Network against Corrupted Labels in Medical Image Segmentation with Domain Shift
• Multi-Source Domain Adaptation with Collaborative Learning for Semantic Segmentation
• Contextual-Relation Consistent Domain Adaptation for Semantic Segmentation
• Improving Semantic Segmentation via Self-Training
• Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach
• CSCL: Critical Semantic-Consistent Learning for Unsupervised Domain Adaptation
• Re-distributing Biased Pseudo Labels for Semi-supervised Semantic Segmentation: A Baseline Investigation
• Domain Adaptation on Semantic Segmentation for Aerial Images
• Contrastive Learning and Self-Training for Unsupervised Domain Adaptation in Semantic Segmentation
• Self-Adaptive Partial Domain Adaptation
• MLAN: Multi-Level Adversarial Network for Domain Adaptive Semantic Segmentation
• Energy-constrained Self-training for Unsupervised Domain Adaptation
• Bridging the Inter-Domain Gap through Low-Level Features for Cross-Modal Medical Image Segmentation
• Global and Local Texture Randomization for Synthetic-to-Real Semantic Segmentation
• Unsupervised Domain Adaptation in Semantic Segmentation: a Review
• Single slice thigh CT muscle group segmentation with domain adaptation and self-training
• Boosting Cross-spectral Unsupervised Domain Adaptation for Thermal Semantic Segmentation
• Cross-View Regularization for Domain Adaptive Panoptic Segmentation
• A Sensor Agnostic Domain Generalization Framework for Leveraging Geospatial Foundation Models: Enhancing Semantic Segmentation viaSynergistic Pseudo-Labeling and Generative Learning
• Test time Adaptation through Perturbation Robustness
• Boundary and Entropy-driven Adversarial Learning for Fundus Image Segmentation
• Unbox the Black-box for the Medical Explainable AI via Multi-modal and Multi-centre Data Fusion: A Mini-Review, Two Showcases and Beyond
• Target Consistency for Domain Adaptation: when Robustness meets Transferability
• CoDEx: Combining Domain Expertise for Spatial Generalization in Satellite Image Analysis
• What is the Added Value of UDA in the VFM Era?
• DyCON: Dynamic Uncertainty-aware Consistency and Contrastive Learning for Semi-supervised Medical Image Segmentation
• On Direct Distribution Matching for Adapting Segmentation Networks
• Adversarial Shape Learning for Building Extraction in VHR Remote Sensing Images
• Adversarial Domain Adaptation With Prototype-Based Normalized Output Conditioner
• WT-BCP: Wavelet Transform based Bidirectional Copy-Paste for Semi-Supervised Medical Image Segmentation
• Multi-Target Domain Adaptation with Collaborative Consistency Learning
• HSACNet: Hierarchical Scale-Aware Consistency Regularized Semi-Supervised Change Detection
• Unsupervised Domain Adaptation Network with Category-Centric Prototype Aligner for Biomedical Image Segmentation
• Synthetic training data generation for deep learning based quality inspection
• FLOSS: Free Lunch in Open-vocabulary Semantic Segmentation
• DUDA: Distilled Unsupervised Domain Adaptation for Lightweight Semantic Segmentation

Related

• Domain Adaptation for Semantic Segmentation with Maximum Squares Loss
• Robust Unsupervised Domain Adaptation by Retaining Confident Entropy via Edge Concatenation
• Adversarial Semi-Supervised Domain Adaptation for Semantic Segmentation: A New Role for Labeled Target Samples
• BiMaL: Bijective Maximum Likelihood Approach to Domain Adaptation in Semantic Scene Segmentation
• Threshold-adaptive Unsupervised Focal Loss for Domain Adaptation of Semantic Segmentation