MoASE++: Mixture of Activation Sparsity Experts with Domain-Adaptive On-policy Distillation for Continual Test Time Adaptation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916021311373312 |
|---|---|
| author | Zhang, Ronyu Cheng, Aosong Dai, Gaole Luo, Yulin Liu, Jiaming Du, Li Yang, Huanrui Wang, Dan Fang, Leyuan Du, Yuan Zhang, Shanghang |
| author_facet | Zhang, Ronyu Cheng, Aosong Dai, Gaole Luo, Yulin Liu, Jiaming Du, Li Yang, Huanrui Wang, Dan Fang, Leyuan Du, Yuan Zhang, Shanghang |
| contents | Continual test-time adaptation adapts a source-pretrained model to non-stationary, unlabeled target streams while retaining past competence, yet texture-biased backbones risk error accumulation and catastrophic forgetting. Drawing inspiration from the process of decoupling shape and texture in the human visual system, we introduce MoASE, a plug-in mixture-of-experts that disentangles domain-agnostic structure from domain-specific texture using Activation Sparsity Experts with Spatial Differentiable Dropout, forming complementary high- and low-activation pathways, while high- and low-rank bottlenecks diversify representations. The Activation Sparsity Gate produces input-adaptive SDD thresholds for precise token selection, and the Domain-Aware Router assigns per-sample expert weights using texture-sensitive cues. To curb confirmation bias on unlabeled streams and stabilize supervision, we then introduce Domain-Adaptive On-Policy Distillation to constitute MoASE++, with an EMA-anchored on-policy reverse KL distillation and an augmentation policy conditioned on entropy and confidence that aligns predictions across the same views and improves the robustness-plasticity balance. Extensive experiments on classification (CIFAR-10/100-C, ImageNet-C) and semantic segmentation (Cityscapes->ACDC) demonstrate consistent state-of-the-art performance, offering a principled, controllable approach to continual adaptation in dynamic visual environments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_17743 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MoASE++: Mixture of Activation Sparsity Experts with Domain-Adaptive On-policy Distillation for Continual Test Time Adaptation Zhang, Ronyu Cheng, Aosong Dai, Gaole Luo, Yulin Liu, Jiaming Du, Li Yang, Huanrui Wang, Dan Fang, Leyuan Du, Yuan Zhang, Shanghang Computer Vision and Pattern Recognition Continual test-time adaptation adapts a source-pretrained model to non-stationary, unlabeled target streams while retaining past competence, yet texture-biased backbones risk error accumulation and catastrophic forgetting. Drawing inspiration from the process of decoupling shape and texture in the human visual system, we introduce MoASE, a plug-in mixture-of-experts that disentangles domain-agnostic structure from domain-specific texture using Activation Sparsity Experts with Spatial Differentiable Dropout, forming complementary high- and low-activation pathways, while high- and low-rank bottlenecks diversify representations. The Activation Sparsity Gate produces input-adaptive SDD thresholds for precise token selection, and the Domain-Aware Router assigns per-sample expert weights using texture-sensitive cues. To curb confirmation bias on unlabeled streams and stabilize supervision, we then introduce Domain-Adaptive On-Policy Distillation to constitute MoASE++, with an EMA-anchored on-policy reverse KL distillation and an augmentation policy conditioned on entropy and confidence that aligns predictions across the same views and improves the robustness-plasticity balance. Extensive experiments on classification (CIFAR-10/100-C, ImageNet-C) and semantic segmentation (Cityscapes->ACDC) demonstrate consistent state-of-the-art performance, offering a principled, controllable approach to continual adaptation in dynamic visual environments. |
| title | MoASE++: Mixture of Activation Sparsity Experts with Domain-Adaptive On-policy Distillation for Continual Test Time Adaptation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.17743 |