Feature-Based Instance Neighbor Discovery: Advanced Stable Test-Time Adaptation in Dynamic World
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916784143073280 |
|---|---|
| author | Jiang, Qinting Ye, Chuyang Wei, Dongyan Wang, Bingli Xue, Yuan Jiang, Jingyan Wang, Zhi |
| author_facet | Jiang, Qinting Ye, Chuyang Wei, Dongyan Wang, Bingli Xue, Yuan Jiang, Jingyan Wang, Zhi |
| contents | Despite progress, deep neural networks still suffer performance declines under distribution shifts between training and test domains, leading to a substantial decrease in Quality of Experience (QoE) for applications. Existing test-time adaptation (TTA) methods are challenged by dynamic, multiple test distributions within batches. We observe that feature distributions across different domains inherently cluster into distinct groups with varying means and variances. This divergence reveals a critical limitation of previous global normalization strategies in TTA, which inevitably distort the original data characteristics. Based on this insight, we propose Feature-based Instance Neighbor Discovery (FIND), which comprises three key components: Layer-wise Feature Disentanglement (LFD), Feature Aware Batch Normalization (FABN) and Selective FABN (S-FABN). LFD stably captures features with similar distributions at each layer by constructing graph structures. While FABN optimally combines source statistics with test-time distribution specific statistics for robust feature representation. Finally, S-FABN determines which layers require feature partitioning and which can remain unified, thereby enhancing inference efficiency. Extensive experiments demonstrate that FIND significantly outperforms existing methods, achieving a 30\% accuracy improvement in dynamic scenarios while maintaining computational efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_06782 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Feature-Based Instance Neighbor Discovery: Advanced Stable Test-Time Adaptation in Dynamic World Jiang, Qinting Ye, Chuyang Wei, Dongyan Wang, Bingli Xue, Yuan Jiang, Jingyan Wang, Zhi Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition Despite progress, deep neural networks still suffer performance declines under distribution shifts between training and test domains, leading to a substantial decrease in Quality of Experience (QoE) for applications. Existing test-time adaptation (TTA) methods are challenged by dynamic, multiple test distributions within batches. We observe that feature distributions across different domains inherently cluster into distinct groups with varying means and variances. This divergence reveals a critical limitation of previous global normalization strategies in TTA, which inevitably distort the original data characteristics. Based on this insight, we propose Feature-based Instance Neighbor Discovery (FIND), which comprises three key components: Layer-wise Feature Disentanglement (LFD), Feature Aware Batch Normalization (FABN) and Selective FABN (S-FABN). LFD stably captures features with similar distributions at each layer by constructing graph structures. While FABN optimally combines source statistics with test-time distribution specific statistics for robust feature representation. Finally, S-FABN determines which layers require feature partitioning and which can remain unified, thereby enhancing inference efficiency. Extensive experiments demonstrate that FIND significantly outperforms existing methods, achieving a 30\% accuracy improvement in dynamic scenarios while maintaining computational efficiency. |
| title | Feature-Based Instance Neighbor Discovery: Advanced Stable Test-Time Adaptation in Dynamic World |
| topic | Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2506.06782 |