TRUST: Test-Time Refinement using Uncertainty-Guided SSM Traverses

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dastani, Sahar, Bahri, Ali, Hakim, Gustavo Adolfo Vargas, Yazdanpanah, Moslem, Noori, Mehrdad, Osowiechi, David, Barbeau, Samuel, Ayed, Ismail Ben, Lombaert, Herve, Desrosiers, Christian
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918488742821888
author Dastani, Sahar
Bahri, Ali
Hakim, Gustavo Adolfo Vargas
Yazdanpanah, Moslem
Noori, Mehrdad
Osowiechi, David
Barbeau, Samuel
Ayed, Ismail Ben
Lombaert, Herve
Desrosiers, Christian
author_facet Dastani, Sahar
Bahri, Ali
Hakim, Gustavo Adolfo Vargas
Yazdanpanah, Moslem
Noori, Mehrdad
Osowiechi, David
Barbeau, Samuel
Ayed, Ismail Ben
Lombaert, Herve
Desrosiers, Christian
contents State Space Models (SSMs) have emerged as efficient alternatives to Vision Transformers (ViTs), with VMamba standing out as a pioneering architecture designed for vision tasks. However, their generalization performance degrades significantly under distribution shifts. To address this limitation, we propose TRUST (Test-Time Refinement using Uncertainty-Guided SSM Traverses), a novel test-time adaptation (TTA) method that leverages diverse traversal permutations to generate multiple causal perspectives of the input image. Model predictions serve as pseudo-labels to guide updates of the Mamba-specific parameters, and the adapted weights are averaged to integrate the learned information across traversal scans. Altogether, TRUST is the first approach that explicitly leverages the unique architectural properties of SSMs for adaptation. Experiments on seven benchmarks show that TRUST consistently improves robustness and outperforms existing TTA methods.
format Preprint
id arxiv_https___arxiv_org_abs_2509_22813
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TRUST: Test-Time Refinement using Uncertainty-Guided SSM Traverses
Dastani, Sahar
Bahri, Ali
Hakim, Gustavo Adolfo Vargas
Yazdanpanah, Moslem
Noori, Mehrdad
Osowiechi, David
Barbeau, Samuel
Ayed, Ismail Ben
Lombaert, Herve
Desrosiers, Christian
Computer Vision and Pattern Recognition
State Space Models (SSMs) have emerged as efficient alternatives to Vision Transformers (ViTs), with VMamba standing out as a pioneering architecture designed for vision tasks. However, their generalization performance degrades significantly under distribution shifts. To address this limitation, we propose TRUST (Test-Time Refinement using Uncertainty-Guided SSM Traverses), a novel test-time adaptation (TTA) method that leverages diverse traversal permutations to generate multiple causal perspectives of the input image. Model predictions serve as pseudo-labels to guide updates of the Mamba-specific parameters, and the adapted weights are averaged to integrate the learned information across traversal scans. Altogether, TRUST is the first approach that explicitly leverages the unique architectural properties of SSMs for adaptation. Experiments on seven benchmarks show that TRUST consistently improves robustness and outperforms existing TTA methods.
title TRUST: Test-Time Refinement using Uncertainty-Guided SSM Traverses
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.22813