Introduces (ε,δ)-Recovery Complexity and (ε,ρ)-TTA Learnability with order-wise matching bounds on recovery time for TTA under gradual and abrupt shifts via a discrete surrogate model.
Beyond model adaptation at test time: A survey
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
abstract
Machine learning algorithms have achieved remarkable success across various disciplines, use cases and applications, under the prevailing assumption that training and test samples are drawn from the same distribution. Consequently, these algorithms struggle and become brittle even when samples in the test distribution start to deviate from the ones observed during training. Domain adaptation and domain generalization have been studied extensively as approaches to address distribution shifts across test and train domains, but each has its limitations. Test-time adaptation, a recently emerging learning paradigm, combines the benefits of domain adaptation and domain generalization by training models only on source data and adapting them to target data during test-time inference. In this survey, we provide a comprehensive and systematic review on test-time adaptation, covering more than 400 recent papers. We structure our review by categorizing existing methods into five distinct categories based on what component of the method is adjusted for test-time adaptation: the model, the inference, the normalization, the sample, or the prompt, providing detailed analysis of each. We further discuss the various preparation and adaptation settings for methods within these categories, offering deeper insights into the effective deployment for the evaluation of distribution shifts and their real-world application in understanding images, video and 3D, as well as modalities beyond vision. We close the survey with an outlook on emerging research opportunities for test-time adaptation.
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
Introduces Dynamic Style Bridging for forward-facilitation continual test-time adaptation by multi-level style injection on pre-generated class proxies to provide stable on-demand supervision under evolving distribution shifts.
An uncertainty-aware test-time adaptation framework improves cross-region spatio-temporal fusion of land surface temperature by updating only the fusion module guided by epistemic uncertainty, land use consistency, and bias correction.
Combines offline behavioral cloning with online Real-Time Recurrent RL fine-tuning on LrcSSM models to adapt autonomous driving policies to distribution shifts, validated in simulation and on a real 1:10-scale robot with event camera.
SAFER enhances robustness of test-time adaptation to adversarial attacks via reliability-guided ensembling of stochastic augmentations while preserving clean performance.
DOME learns sample-specific domain variables from sparse supervision via vision-language models and a sparse domain bank to improve test-time adaptation performance.
Entropy minimization amplifies prediction bias from merged feature clusters under distribution shifts, and DSBR mitigates collapse by equalizing predicted class contributions to the unsupervised loss.
Ramen enables robust test-time adaptation of vision-language models under mixed-domain shifts by actively selecting domain-consistent and prediction-balanced samples via an embedding-gradient cache.
A hierarchical taxonomy and comparative review of continual test-time adaptation methods that adapt vision models online to non-stationary unlabeled streams while fighting forgetting and error accumulation.
Edge AI systems require ongoing adaptation to evolving data and constraints to avoid violating budgets or losing reliability, formalized via an Agent-System-Environment lens that defines ten future research challenges.
Presents a state-space modeling probabilistic framework for online test-time adaptation under distributional shift.
citing papers explorer
-
On the Learnability of Test-Time Adaptation: A Recovery Complexity Perspective
Introduces (ε,δ)-Recovery Complexity and (ε,ρ)-TTA Learnability with order-wise matching bounds on recovery time for TTA under gradual and abrupt shifts via a discrete surrogate model.
-
Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging
Introduces Dynamic Style Bridging for forward-facilitation continual test-time adaptation by multi-level style injection on pre-generated class proxies to provide stable on-demand supervision under evolving distribution shifts.
-
Uncertainty-Aware Test-Time Adaptation for Cross-Region Spatio-Temporal Fusion of Land Surface Temperature
An uncertainty-aware test-time adaptation framework improves cross-region spatio-temporal fusion of land surface temperature by updating only the fusion module guided by epistemic uncertainty, land use consistency, and bias correction.
-
Adaptive Control in Autonomous Driving via Real-Time Recurrent RL
Combines offline behavioral cloning with online Real-Time Recurrent RL fine-tuning on LrcSSM models to adapt autonomous driving policies to distribution shifts, validated in simulation and on a real 1:10-scale robot with event camera.
-
Reliability-Guided Adaptive Ensembling for Robust Test-Time Adaptation
SAFER enhances robustness of test-time adaptation to adversarial attacks via reliability-guided ensembling of stochastic augmentations while preserving clean performance.
-
DOME: Learning Transferable Domain Variables from Sparse Supervision for Test-Time Adaptation
DOME learns sample-specific domain variables from sparse supervision via vision-language models and a sparse domain bank to improve test-time adaptation performance.
-
Entropy Minimization without Model Collapse: Mitigating Prediction Bias in Medical Imaging
Entropy minimization amplifies prediction bias from merged feature clusters under distribution shifts, and DSBR mitigates collapse by equalizing predicted class contributions to the unsupervised loss.
-
Ramen: Robust Test-Time Adaptation of Vision-Language Models with Active Sample Selection
Ramen enables robust test-time adaptation of vision-language models under mixed-domain shifts by actively selecting domain-consistent and prediction-balanced samples via an embedding-gradient cache.
-
Continual Test-Time Adaptation in Computer Vision: Methods, Benchmarks, and Future Directions
A hierarchical taxonomy and comparative review of continual test-time adaptation methods that adapt vision models online to non-stationary unlabeled streams while fighting forgetting and error accumulation.
-
Position Paper: From Edge AI to Adaptive Edge AI
Edge AI systems require ongoing adaptation to evolving data and constraints to avoid violating budgets or losing reliability, formalized via an Agent-System-Environment lens that defines ten future research challenges.
-
A probabilistic framework for online test-time adaptation
Presents a state-space modeling probabilistic framework for online test-time adaptation under distributional shift.
- EmbodiTTA: Resource-Efficient Test-Time Adaptation for Embodied Visual Systems