Test-Time Adaptation via Dual Distillation for Videos Under Severe Distribution Shifts

• URL: http://arxiv.org/abs/2607.24611v1
• Date: Mon, 27 Jul 2026 16:12:11 GMT
• Title: Test-Time Adaptation via Dual Distillation for Videos Under Severe Distribution Shifts
• Abstract summary: Deep learning models experience severe performance degradation when applied to real-world scenarios due to unanticipated distribution shifts.<n>Test-Time Adaptation (TTA) attempts to solve this problem by using unlabeled data from the target domain to dynamically adapt to the test distribution at inference time.<n>We propose Test-time Adaptation via Dual Distillation (TADD), an online adaptation framework that relies on a lightweight projection adapter to bridge the domain gap.
• Score: 1.5539424546888112
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deep learning models have achieved state-of-the-art performance in several computer vision tasks. However, they experience severe performance degradation when applied to real-world scenarios due to unanticipated distribution shifts. Test-Time Adaptation (TTA) attempts to solve this problem by using unlabeled data from the target domain to dynamically adapt to the test distribution at inference time, without access to the source data. However, TTA remains a challenging problem when adapting to continuous, temporally correlated data, such as videos, and in scenarios where the target domain contains severe domain shifts. For this reason, few works in the literature explore TTA for videos under such extreme conditions. To overcome these limitations, we propose Test-time Adaptation via Dual Distillation (TADD), an online adaptation framework that relies on a lightweight projection adapter to bridge the domain gap. The adapter module is pre-trained on the source domain and then adapted to the target using our proposed complementary losses: (i) zero-shot distillation, which encourages alignment with the domain-agnostic features from a pre-trained vision-language model (VLM); and (ii) target distillation, which retains the source domain discriminative knowledge encoded in the pre-trained adapter. Built upon a frozen CLIP backbone, our method introduces this lightweight projection adapter as the sole updatable component during inference. We conducted extensive evaluations on three well-known video action recognition benchmarks: UCF-HMDB, Daily-DA, and Sports-DA. Our experiments in the closed-set scenario demonstrate that our method consistently outperforms state-of-the-art TTA baselines. Notably, our TTA approach improves upon previous methods by up to +3.81% on UCF-HMDB, +2.63% on Daily-DA, and +3.03% on Sports-DA.

Related papers

• TestMate: Test-Time Domain Adaptation Aided by Lightweight Vision Foundation Model [21.331953401986322]
  Test-Time Domain Adaptation (TTDA) aims to adapt Deep Neural Networks to distribution shifts using only streaming, unlabeled test data in real time.<n>We introduce TestMate, a novel, real-time, and backpropagation-free TTDA framework that overcomes these issues.
  arXiv  Detail & Related papers  (2026-07-04T10:32:15Z)
• Enhancing Test Time Adaptation with Few-shot Guidance [62.49199492255226]
  Deep neural networks often encounter significant performance drops while facing with domain shifts between training (source) and test (target) data.<n>Test Time Adaptation (TTA) methods have been proposed to adapt pre-trained source model to handle out-of-distribution streaming target data.<n>We develop Few-Shot Test Time Adaptation (FS-TTA), a novel and practical setting that utilizes a few-shot support set on top of TTA.
  arXiv  Detail & Related papers  (2024-09-02T15:50:48Z)
• Learn from the Learnt: Source-Free Active Domain Adaptation via Contrastive Sampling and Visual Persistence [60.37934652213881]
  Domain Adaptation (DA) facilitates knowledge transfer from a source domain to a related target domain. This paper investigates a practical DA paradigm, namely Source data-Free Active Domain Adaptation (SFADA), where source data becomes inaccessible during adaptation. We present learn from the learnt (LFTL), a novel paradigm for SFADA to leverage the learnt knowledge from the source pretrained model and actively iterated models without extra overhead.
  arXiv  Detail & Related papers  (2024-07-26T17:51:58Z)
• Test-time adaptation for geospatial point cloud semantic segmentation with distinct domain shifts [6.80671668491958]
  Test-time adaptation (TTA) allows direct adaptation of a pre-trained model to unlabeled data during inference stage without access to source data or additional training. We propose three domain shift paradigms: photogrammetric to airborne LiDAR, airborne to mobile LiDAR, and synthetic to mobile laser scanning. Experimental results show our method improves classification accuracy by up to 20% mIoU, outperforming other methods.
  arXiv  Detail & Related papers  (2024-07-08T15:40:28Z)
• Spatio-Temporal Pixel-Level Contrastive Learning-based Source-Free Domain Adaptation for Video Semantic Segmentation [117.39092621796753]
  Source Domain Adaptation (SFDA) setup aims to adapt a source-trained model to the target domain without accessing source data. A novel method that takes full advantage of correlations oftemporal-information to tackle the absence of source data is proposed. Experiments show that PixelL achieves un-of-the-art performance on benchmarks compared to current UDA and SFDA approaches.
  arXiv  Detail & Related papers  (2023-03-25T05:06:23Z)
• Confidence Attention and Generalization Enhanced Distillation for Continuous Video Domain Adaptation [62.458968086881555]
  Continuous Video Domain Adaptation (CVDA) is a scenario where a source model is required to adapt to a series of individually available changing target domains. We propose a Confidence-Attentive network with geneRalization enhanced self-knowledge disTillation (CART) to address the challenge in CVDA.
  arXiv  Detail & Related papers  (2023-03-18T16:40:10Z)
• Domain Alignment Meets Fully Test-Time Adaptation [24.546705919244936]
  A foundational requirement of a deployed ML model is to generalize to data drawn from a testing distribution that is different from training. In this paper, we focus on a challenging variant of this problem, where access to the original source data is restricted. We propose a new approach, CATTAn, that bridges UDA and FTTA, by relaxing the need to access entire source data.
  arXiv  Detail & Related papers  (2022-07-09T03:17:19Z)
• Back to the Source: Diffusion-Driven Test-Time Adaptation [77.4229736436935]
  Test-time adaptation harnesses test inputs to improve accuracy of a model trained on source data when tested on shifted target data. We instead update the target data, by projecting all test inputs toward the source domain with a generative diffusion model.
  arXiv  Detail & Related papers  (2022-07-07T17:14:10Z)
• CAFA: Class-Aware Feature Alignment for Test-Time Adaptation [50.26963784271912]
  Test-time adaptation (TTA) aims to address this challenge by adapting a model to unlabeled data at test time. We propose a simple yet effective feature alignment loss, termed as Class-Aware Feature Alignment (CAFA), which simultaneously encourages a model to learn target representations in a class-discriminative manner.
  arXiv  Detail & Related papers  (2022-06-01T03:02:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.