Star the repo to show support! β¨
This repository systematically organizes research advancements in CTTA, primarily featuring published articles in this rapidly evolving field.
Deep learning models often struggle when deployed in real-world settings with non-stationary environments. Developing techniques for continual adaptation during unsupervised deployment is essential for reliable machine perception systems.
This repo serves as a community-driven hub for:
- π Tracking progress: Documenting the evolution of CTTA methods through a chronological timeline.
- π Organizing approaches: Categorizing methods by technical approaches.
- π Fostering collaboration: Encouraging open discussions and contributions.
π Timeline
Sorted chronologically (newest first).
Year | Venue | Paper Title | Code/Project |
---|---|---|---|
2025 | CVPR | Maintaining Consistent Inter-Class Topology in Continual Test-Time Adaptation | Code |
2025 | CVPR | Unsupervised Continual Domain Shift Learning with Multi-Prototype Modeling | N/A |
Year | Venue | Paper Title | Code/Project |
---|---|---|---|
2022 | CVPR | Continual Test-Time Domain Adaptation | Code |
2022 | ICML | Efficient Test-Time Model Adaptation without Forgetting | Code |
π§° Technical Approaches
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2022 | CVPR | Random | Continual Test-Time Domain Adaptation | Code |
2022 | ICML | Data-driven, Fisher | Efficient Test-Time Model Adaptation without Forgetting | Code |
2023 | CVPR | Data-driven, Fisher | A Probabilistic Framework for Lifelong Test-Time Adaptation | Code |
2023 | NeurIPS | Periodical | RDumb: A simple approach that questions our progress in continual test-time adaptation | Code |
2024 | WACV | Peak Detection | Effective Restoration of Source Knowledge in Continual Test Time Adaptation | N/A |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2022 | ICML | Entropy | Efficient Test-Time Model Adaptation without Forgetting | Code |
2023 | IJCAI | Self-adaptive Thresholds | Exploring Safety Supervision for Continual Test-time Domain Adaptation | N/A |
2024 | IJCAI | Self-adaptive Thresholds | Navigating Continual Test-time Adaptation with Symbiosis Knowledge | N/A |
2024 | CVPR | Self-adaptive Thresholds | A Versatile Framework for Continual Test-Time Domain Adaptation: Balancing Discriminability and Generalizability | N/A |
2024 | WACV | Dynamic Thresholds | Continual Test-time Domain Adaptation via Dynamic Sample Selection | N/A |
2024 | CVPR | Entropy | BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptation | Code |
2024 | ECCV | Entropy | Reshaping the Online Data Buffering and Organizing Mechanism for Continual Test-Time Adaptation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2023 | CVPR | Source Prototypes | Robust Mean Teacher for Continual and Gradual Test-Time Adaptation | Code |
2023 | TMLR | Source Prototypes | SANTA: Source Anchoring Network and Target Alignment for Continual Test Time Adaptation | Code |
2023 | IJCAI | Reweighted | Exploring Safety Supervision for Continual Test-time Domain Adaptation | N/A |
2024 | IJCAI | Reweighted | Navigating Continual Test-time Adaptation with Symbiosis Knowledge | N/A |
2025 | CVPR | Multi-granularity | Maintaining Consistent Inter-Class Topology in Continual Test-Time Adaptation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2023 | CVPR | Source Samples | Robust Mean Teacher for Continual and Gradual Test-Time Adaptation | Code |
2024 | BMVC | Source Samples | AR-TTA: A Simple Method for Real-World Continual Test-Time Adaptation | N/A |
2024 | ICASSP | Diversity | Diversity-aware Buffer for Coping with Temporally Correlated Data Streams in Online Test-time Adaptation | N/A |
2024 | ECCV | Target Samples | Reshaping the Online Data Buffering and Organizing Mechanism for Continual Test-Time Adaptation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2023 | AAAI | Input Prompts | Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptation | N/A |
2023 | CVPR | Meta Nets | EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularization | Community Implementation |
2024 | ICLR | Adapters | ViDA: Homeostatic Visual Domain Adapter for Continual Test Time Adaptation | Code |
2024 | ICRA | Selective | Distribution-Aware Continual Test Time Adaptation for Semantic Segmentation | Code |
2024 | CVPR | MoE with Adapters | BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptation | Code |
2024 | CVPR | Low-frequency Prompt | Each Test Image Deserves A Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Segmentation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2024 | CVPR | HOG Recons | Continual-MAE: Adaptive Distribution Masked Autoencoders for Continual Test-Time Adaptation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2023 | CVPR | Mean Teacher | Robust Mean Teacher for Continual and Gradual Test-Time Adaptation | Code |
2023 | AAAI | Input Prompts | Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptation | N/A |
2023 | CVPR | Meta Nets | EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularization | Community Implementation |
2024 | ICLR | Adapters | ViDA: Homeostatic Visual Domain Adapter for Continual Test Time Adaptation | Code |
2024 | CVPR | Domain Augmentation, Routers, Adapters | BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptation | Code |
Year | Venue | Key Words | Paper Title | Code/Project |
---|---|---|---|---|
2024 | ECCV | Domain-invariant Inter-class Relation | Reshaping the Online Data Buffering and Organizing Mechanism for Continual Test-Time Adaptation | Code |
2025 | CVPR | Uniform Topological Relation | Maintaining Consistent Inter-Class Topology in Continual Test-Time Adaptation | Code |
π Related Resources
Feel free to contact me (visiongroupzzl@gmail.com) if you find any interesting paper is missing.
Acknowledgments: This repo builds on the efforts of the open-source community. We appreciate everyone's contributions.