---
title: "SAME — paper by Yang et al. (2026) — summary and ideas"
description: "The authors describe this as the first study of test-time adaptation (TTA) for SLT. 41 references."
authors: ["Lujia Yang", "Weicai Yan", "Yongbo He", "Qifei Zhang", "Tao Jin", "Jinshan Zhang", "Meng Xi", "Jianwei Yin"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.973/"
doi: "10.18653/v1/2026.acl-long.973"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/yang-2026-same-signer-aware-mixture-experts/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/yang-2026-same-signer-aware-mixture-experts/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/48/) / SAME: Signer-Aware Mixture-of-Experts for Test-Time Adaptation in Sign Language Translation

# SAME: Signer-Aware Mixture-of-Experts for Test-Time Adaptation in Sign Language Translation

**Authors:** Lujia Yang, [Weicai Yan](https://md.research.elamur.ai/free/en/authors/weicai-yan/), [Yongbo He](https://md.research.elamur.ai/free/en/authors/yongbo-he/), [Qifei Zhang](https://md.research.elamur.ai/free/en/authors/qifei-zhang/), [Tao Jin](https://md.research.elamur.ai/free/en/authors/tao-jin/), Jinshan Zhang, [Meng Xi](https://md.research.elamur.ai/free/en/authors/meng-xi/) and [Jianwei Yin](https://md.research.elamur.ai/free/en/authors/jianwei-yin/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-long.973/) · [PDF](https://aclanthology.org/2026.acl-long.973.pdf) · [DOI 10.18653/v1/2026.acl-long.973](https://doi.org/10.18653/v1/2026.acl-long.973) · [OpenAlex](https://openalex.org/W7166824988)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Sign language translation (SLT) systems often lose accuracy in deployment when signers, lighting or backgrounds differ from the training data. Labeled target data is rarely available for fine-tuning, and many unsupervised adaptation methods depend on batch statistics or slow adaptation. The authors describe this as the first study of test-time adaptation (TTA) for SLT. In their setting, unlabeled videos arrive one by one, and the model is reset and briefly updated for each sample.

Their method, SAME, freezes a pretrained SLT encoder-decoder and adds small Mixture-of-Experts modules after each encoder layer. Each module contains LoRA experts and a pass-through option. A top-k router chooses experts frame by frame, using both the visual features and signer embeddings. Before deployment, the modules are initialized on a small labeled source subset with an orthogonality-based diversity loss to prevent expert collapse. At test time, the model runs a few gradient steps on a joint objective with three parts: lowering the entropy of its predictions, reducing confusion between classes, and learning from its own pseudo-labels.

On Phoenix-2014T, CSL-Daily and How2Sign, SAME reportedly beats prior TTA methods and improves several SLT backbones. This makes on-the-fly personalization feasible for assistive translation.

## Key points
- Evaluation covers three benchmarks: Phoenix-2014T, CSL-Daily and How2Sign. Baseline TTA methods include TENT, SAR and BeCoTTA.
- Ablations show that signer-aware routing contributes most. MoE initialization is second, while frame-wise routing and the pass-through branch give smaller but consistent gains. Combining EM, MCC and PLS works better than any single loss.
- The best hyperparameters were 4 experts, LoRA rank 8 and top-1 routing. In the top-k ablation, top-1 reached BLEU-1 37.61 and BLEU-4 12.77, versus 11.68 BLEU-4 for top-2.
- Cost on Phoenix-2014T: SAME needs 1.62 s/sample and 64.47 GFLOPs. The other methods measured:
  - TENT: 1.35 s, 50.81 GFLOPs
  - SAR: 1.92 s, 79.78 GFLOPs
  - BeCoTTA: 2.75 s, 110.13 GFLOPs
  - Unadapted source model: 0.59 s
- Visualizations show that different signers use different expert subsets and that experts change across a video's temporal segments. The stated limitations are added inference latency and sensitivity to noisy input streams.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Zhejiang University](https://md.research.elamur.ai/free/en/universities/zhejiang-university/) — University, Hangzhou, China · 224 works · 13,587 citations

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works · [🖱️ Human-Computer Interaction](https://md.research.elamur.ai/free/en/shelves/computer-science/human-computer-interaction/) · 160 works

## Cite

Cite as: Lujia Yang, Weicai Yan, Yongbo He, Qifei Zhang, Tao Jin, Jinshan Zhang, Meng Xi and Jianwei Yin. 2026. SAME: Signer-Aware Mixture-of-Experts for Test-Time Adaptation in Sign Language Translation. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{yang2026same,
  title = {{SAME: Signer-Aware Mixture-of-Experts for Test-Time Adaptation in Sign Language Translation}},
  author = {Lujia Yang, Weicai Yan, Yongbo He, Qifei Zhang, Tao Jin, Jinshan Zhang, Meng Xi and Jianwei Yin},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.973},
  url = {https://doi.org/10.18653/v1/2026.acl-long.973}
}
```

- [← previous SAME: Safety-Aware Model Editing Guided by Safety Transformation](https://md.research.elamur.ai/free/en/papers/wang-2026-same-safety-aware-model-editing/)
- [next → SAMem: State-Aware Memory as a Fine-Grained Memory for LLM Agents in Decision-Making](https://md.research.elamur.ai/free/en/papers/wang-2026-samem-state-aware-memory-as/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/48/)

## References

41 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Necati Cihan Camgoz, Simon Hadfield, Oscar Koller, Hermann Ney, and Richard Bowden. 2018. Neural sign language translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 7784–7793. [find](https://scholar.google.com/scholar?q=Necati+Cihan+Camgoz%2C+Simon+Hadfield%2C+Oscar+Koller%2C+Hermann+Ney%2C+and+Richard+Bowden%2E+2018%2E+Neural+sign+language+translation%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+Vision+and+Pattern+Rec)
2. Necati Cihan Camgoz, Oscar Koller, Simon Hadfield, and Richard Bowden. 2020. Sign language transformers: Joint end-to-end sign language recognition and translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10023–10033. [find](https://scholar.google.com/scholar?q=Necati+Cihan+Camgoz%2C+Oscar+Koller%2C+Simon+Hadfield%2C+and+Richard+Bowden%2E+2020%2E+Sign+language+transformers%3A+Joint+end%2Dto%2Dend+sign+language+recognition+and+translation%2E+In+Proceedings+of+the+IEEE%2FCVF+Conf)
3. Shaoxiang Chen, Zequn Jie, and Lin Ma. 2024. LLaVA-MoLE: Sparse mixture of LoRA experts for mitigating data conflicts in instruction finetuning MLLMs. arXiv preprint arXiv:2401.16160. — [arXiv:2401.16160](https://arxiv.org/abs/2401.16160) [find](https://scholar.google.com/scholar?q=Shaoxiang+Chen%2C+Zequn+Jie%2C+and+Lin+Ma%2E+2024%2E+LLaVA%2DMoLE%3A+Sparse+mixture+of+LoRA+experts+for+mitigating+data+conflicts+in+instruction+finetuning+MLLMs%2E+arXiv+preprint+arXiv%3A2401%2E16160%2E+%E2%80%94+%5BarXiv%3A2401%2E16)
4. Yutong Chen, Fangyun Wei, Xiao Sun, Zhirong Wu, and Stephen Lin. 2022a. A simple multi-modality transfer learning baseline for sign language translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 5120–5130. [find](https://scholar.google.com/scholar?q=Yutong+Chen%2C+Fangyun+Wei%2C+Xiao+Sun%2C+Zhirong+Wu%2C+and+Stephen+Lin%2E+2022a%2E+A+simple+multi%2Dmodality+transfer+learning+baseline+for+sign+language+translation%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+C)
5. Yutong Chen, Ronglai Zuo, Fangyun Wei, Yu Wu, Shujie Liu, and Brian Mak. 2022b. Two-stream network for sign language recognition and translation. Advances in Neural Information Processing Systems, 35:17043–17056. [find](https://scholar.google.com/scholar?q=Yutong+Chen%2C+Ronglai+Zuo%2C+Fangyun+Wei%2C+Yu+Wu%2C+Shujie+Liu%2C+and+Brian+Mak%2E+2022b%2E+Two%2Dstream+network+for+sign+language+recognition+and+translation%2E+Advances+in+Neural+Information+Processing+Systems%2C+35%3A)
6. Vasily Dimitri, Barbara Regina, and Magdolna Alfonz. 2025. A survey on mixture of experts: Advancements, challenges, and future directions. Authorea Preprints. [find](https://scholar.google.com/scholar?q=Vasily+Dimitri%2C+Barbara+Regina%2C+and+Magdolna+Alfonz%2E+2025%2E+A+survey+on+mixture+of+experts%3A+Advancements%2C+challenges%2C+and+future+directions%2E+Authorea+Preprints%2E)
7. Hao Dong, Eleni Chatzi, and Olga Fink. 2025. Towards robust multimodal open-set test-time adaptation via adaptive entropy-aware optimization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Hao+Dong%2C+Eleni+Chatzi%2C+and+Olga+Fink%2E+2025%2E+Towards+robust+multimodal+open%2Dset+test%2Dtime+adaptation+via+adaptive+entropy%2Daware+optimization%2E+In+International+Conference+on+Learning+Representations%2E)
8. Amanda Duarte, Shruti Palaskar, Lucas Ventura, Deepti Ghadiyaram, Kenneth DeHaan, Florian Metze, Jordi Torres, and Xavier Giro-i Nieto. 2021. How2Sign: A large-scale multimodal dataset for continuous american sign language. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec… [find](https://scholar.google.com/scholar?q=Amanda+Duarte%2C+Shruti+Palaskar%2C+Lucas+Ventura%2C+Deepti+Ghadiyaram%2C+Kenneth+DeHaan%2C+Florian+Metze%2C+Jordi+Torres%2C+and+Xavier+Giro%2Di+Nieto%2E+2021%2E+How2Sign%3A+A+large%2Dscale+multimodal+dataset+for+continuous+)
9. Jia Gong, Lin Geng Foo, Yixuan He, Hossein Rahmani, and Jun Liu. 2024. LLMs are good sign language translators. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 18362–18372. [find](https://scholar.google.com/scholar?q=Jia+Gong%2C+Lin+Geng+Foo%2C+Yixuan+He%2C+Hossein+Rahmani%2C+and+Jun+Liu%2E+2024%2E+LLMs+are+good+sign+language+translators%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+Vision+and+Pattern+Recognition+%28CV)
10. Taesik Gong, Yewon Kim, Taeckyung Lee, Sorn Chot-tananurak, and Sung-Ju Lee. 2023. SoTTA: Robust test-time adaptation on noisy data streams. Advances in Neural Information Processing Systems, 36:14070– 14093. [find](https://scholar.google.com/scholar?q=Taesik+Gong%2C+Yewon+Kim%2C+Taeckyung+Lee%2C+Sorn+Chot%2Dtananurak%2C+and+Sung%2DJu+Lee%2E+2023%2E+SoTTA%3A+Robust+test%2Dtime+adaptation+on+noisy+data+streams%2E+Advances+in+Neural+Information+Processing+Systems%2C+36%3A14070)
11. Yongbo He, Zirun Guo, and Tao Jin. 2026. Decoupling stability and plasticity for multi-modal test-time adaptation. arXiv preprint arXiv:2603.00574. — [arXiv:2603.00574](https://arxiv.org/abs/2603.00574) [find](https://scholar.google.com/scholar?q=Yongbo+He%2C+Zirun+Guo%2C+and+Tao+Jin%2E+2026%2E+Decoupling+stability+and+plasticity+for+multi%2Dmodal+test%2Dtime+adaptation%2E+arXiv+preprint+arXiv%3A2603%2E00574%2E+%E2%80%94+%5BarXiv%3A2603%2E00574%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2603%2E0057)
12. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. [LoRA: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). In International Conference on Learning Representations.
13. Robert A. Jacobs, Michael I. Jordan, Steven J. Nowlan, and Geoffrey E. Hinton. 1991. Adaptive mixtures of local experts. Neural Computation, 3(1):79–87. [find](https://scholar.google.com/scholar?q=Robert+A%2E+Jacobs%2C+Michael+I%2E+Jordan%2C+Steven+J%2E+Nowlan%2C+and+Geoffrey+E%2E+Hinton%2E+1991%2E+Adaptive+mixtures+of+local+experts%2E+Neural+Computation%2C+3%281%29%3A79%E2%80%9387%2E)
14. Michael I. Jordan and Robert A. Jacobs. 1994. Hierarchical mixtures of experts and the EM algorithm. Neural Computation, 6(2):181–214. [find](https://scholar.google.com/scholar?q=Michael+I%2E+Jordan+and+Robert+A%2E+Jacobs%2E+1994%2E+Hierarchical+mixtures+of+experts+and+the+EM+algorithm%2E+Neural+Computation%2C+6%282%29%3A181%E2%80%93214%2E)
15. Daeun Lee, Jaehong Yoon, and Sung Ju Hwang. 2024. BECoTTA: Input-dependent online blending of experts for continual test-time adaptation. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pages 27072–27093. PMLR. [find](https://scholar.google.com/scholar?q=Daeun+Lee%2C+Jaehong+Yoon%2C+and+Sung+Ju+Hwang%2E+2024%2E+BECoTTA%3A+Input%2Ddependent+online+blending+of+experts+for+continual+test%2Dtime+adaptation%2E+In+Proceedings+of+the+41st+International+Conference+on+Machine)
16. Zecheng Li, Wengang Zhou, Weichao Zhao, Kepeng Wu, Hezhen Hu, and Houqiang Li. 2025. Uni-Sign: Toward unified sign language understanding at scale. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Zecheng+Li%2C+Wengang+Zhou%2C+Weichao+Zhao%2C+Kepeng+Wu%2C+Hezhen+Hu%2C+and+Houqiang+Li%2E+2025%2E+Uni%2DSign%3A+Toward+unified+sign+language+understanding+at+scale%2E+In+International+Conference+on+Learning+Representati)
17. Guan-Ting Lin, Wei Ping Huang, and Hung-yi Lee. 2024. Continual test-time adaptation for end-to-end speech recognition on noisy speech. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 20003–20015, Miami, Florida, USA. [find](https://scholar.google.com/scholar?q=Guan%2DTing+Lin%2C+Wei+Ping+Huang%2C+and+Hung%2Dyi+Lee%2E+2024%2E+Continual+test%2Dtime+adaptation+for+end%2Dto%2Dend+speech+recognition+on+noisy+speech%2E+In+Proceedings+of+the+2024+Conference+on+Empirical+Methods+in+Na)
18. Guan-Ting Lin, Shang-Wen Li, and Hung-yi Lee. 2022. Listen, adapt, better WER: Source-free single-utterance test-time adaptation for automatic speech recognition. In Proc. [find](https://scholar.google.com/scholar?q=Guan%2DTing+Lin%2C+Shang%2DWen+Li%2C+and+Hung%2Dyi+Lee%2E+2022%2E+Listen%2C+adapt%2C+better+WER%3A+Source%2Dfree+single%2Dutterance+test%2Dtime+adaptation+for+automatic+speech+recognition%2E+In+Proc%2E)
19. Interspeech 2022, pages 2198– 2202. [find](https://scholar.google.com/scholar?q=Interspeech+2022%2C+pages+2198%E2%80%93+2202%2E)
20. Kezhou Lin, Xiaohan Wang, Linchao Zhu, Ke Sun, Bang Zhang, and Yi Yang. 2023. Gloss-free end-to-end sign language translation. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 12904–12916, Toronto, Canada. Association for Compu… [find](https://scholar.google.com/scholar?q=Kezhou+Lin%2C+Xiaohan+Wang%2C+Linchao+Zhu%2C+Ke+Sun%2C+Bang+Zhang%2C+and+Yi+Yang%2E+2023%2E+Gloss%2Dfree+end%2Dto%2Dend+sign+language+translation%2E+In+Proceedings+of+the+61st+Annual+Meeting+of+the+Association+for+Computat)
21. Shuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Yaofo Chen, Shijian Zheng, Peilin Zhao, and Mingkui Tan. 2022. Efficient test-time model adaptation without forgetting. In International conference on machine learning, pages 16888–16905. PMLR. [find](https://scholar.google.com/scholar?q=Shuaicheng+Niu%2C+Jiaxiang+Wu%2C+Yifan+Zhang%2C+Yaofo+Chen%2C+Shijian+Zheng%2C+Peilin+Zhao%2C+and+Mingkui+Tan%2E+2022%2E+Efficient+test%2Dtime+model+adaptation+without+forgetting%2E+In+International+conference+on+machine)
22. Shuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Zhiquan Wen, Yaofo Chen, Peilin Zhao, and Mingkui Tan. 2023. Towards stable test-time adaptation in dynamic wild world. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Shuaicheng+Niu%2C+Jiaxiang+Wu%2C+Yifan+Zhang%2C+Zhiquan+Wen%2C+Yaofo+Chen%2C+Peilin+Zhao%2C+and+Mingkui+Tan%2E+2023%2E+Towards+stable+test%2Dtime+adaptation+in+dynamic+wild+world%2E+In+International+Conference+on+Learnin)
23. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. [BLEU: a method for automatic evaluation of machine translation](https://md.research.elamur.ai/free/en/papers/papineni-2002-bleu-method-automatic-evaluation-machine/). In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318, Philadelphia, Pennsylvania, USA. Association for Computational…
24. Samyam Rajbhandari, Conglong Li, Zhewei Yao, Minjia Zhang, Reza Yazdani Aminabadi, Ammar Ahmad Awan, Jeff Rasley, and Yuxiong He. 2022. [DeepSpeed-MoE: Advancing mixture-of-experts inference and training to power next-generation AI scale](https://md.research.elamur.ai/free/en/papers/rajbhandari-2022-deepspeed-moe-advancing-mixture-experts/). In Proceedings of the 39th International Conference on Machine…
25. Leonardo Rossi, Vittorio Bernuzzi, Tomaso Fontanini, Massimo Bertozzi, and Andrea Prati. 2025. Swin2-MoSE: A new single image supersolution model for remote sensing. IET Image Processing, 19(1):e13303. [find](https://scholar.google.com/scholar?q=Leonardo+Rossi%2C+Vittorio+Bernuzzi%2C+Tomaso+Fontanini%2C+Massimo+Bertozzi%2C+and+Andrea+Prati%2E+2025%2E+Swin2%2DMoSE%3A+A+new+single+image+supersolution+model+for+remote+sensing%2E+IET+Image+Processing%2C+19%281%29%3Ae13303)
26. Jin Shin and Hyun Kim. 2024. L-TTA: Lightweight test-time adaptation using a versatile stem layer. In Advances in Neural Information Processing Systems, volume 37, pages 39325–39349. [find](https://scholar.google.com/scholar?q=Jin+Shin+and+Hyun+Kim%2E+2024%2E+L%2DTTA%3A+Lightweight+test%2Dtime+adaptation+using+a+versatile+stem+layer%2E+In+Advances+in+Neural+Information+Processing+Systems%2C+volume+37%2C+pages+39325%E2%80%9339349%2E)
27. Laia Tarrés, Gerard I. Gállego, Amanda Duarte, Jordi Torres, and Xavier Giró-i Nieto. 2023. Sign language translation from instructional videos. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pages 5625–5635. [find](https://scholar.google.com/scholar?q=Laia+Tarr%C3%A9s%2C+Gerard+I%2E+G%C3%A1llego%2C+Amanda+Duarte%2C+Jordi+Torres%2C+and+Xavier+Gir%C3%B3%2Di+Nieto%2E+2023%2E+Sign+language+translation+from+instructional+videos%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+V)
28. Mathurin Videau, Alessandro Leite, Marc Schoenauer, and Olivier Teytaud. 2024. Mixture of experts in image classification: What’s the sweet spot? arXiv preprint arXiv:2411.18322. — [arXiv:2411.18322](https://arxiv.org/abs/2411.18322) [find](https://scholar.google.com/scholar?q=Mathurin+Videau%2C+Alessandro+Leite%2C+Marc+Schoenauer%2C+and+Olivier+Teytaud%2E+2024%2E+Mixture+of+experts+in+image+classification%3A+What%E2%80%99s+the+sweet+spot%3F+arXiv+preprint+arXiv%3A2411%2E18322%2E+%E2%80%94+%5BarXiv%3A2411%2E18322%5D%28)
29. Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Olshausen, and Trevor Darrell. 2021. Tent: Fully test-time adaptation by entropy minimization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Dequan+Wang%2C+Evan+Shelhamer%2C+Shaoteng+Liu%2C+Bruno+Olshausen%2C+and+Trevor+Darrell%2E+2021%2E+Tent%3A+Fully+test%2Dtime+adaptation+by+entropy+minimization%2E+In+International+Conference+on+Learning+Representations%2E)
30. Qin Wang, Olga Fink, Luc Van Gool, and Dengxin Dai. 2022. Continual test-time domain adaptation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 7201–7211. [find](https://scholar.google.com/scholar?q=Qin+Wang%2C+Olga+Fink%2C+Luc+Van+Gool%2C+and+Dengxin+Dai%2E+2022%2E+Continual+test%2Dtime+domain+adaptation%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+Vision+and+Pattern+Recognition+%28CVPR%29%2C+pages+7201)
31. Zixin Wang, Yadan Luo, Liang Zheng, Zhuoxiao Chen, Sen Wang, and Zi Huang. 2025. In search of lost online test-time adaptation: A survey. International Journal of Computer Vision, 133(3):1106–1139. [find](https://scholar.google.com/scholar?q=Zixin+Wang%2C+Yadan+Luo%2C+Liang+Zheng%2C+Zhuoxiao+Chen%2C+Sen+Wang%2C+and+Zi+Huang%2E+2025%2E+In+search+of+lost+online+test%2Dtime+adaptation%3A+A+survey%2E+International+Journal+of+Computer+Vision%2C+133%283%29%3A1106%E2%80%931139%2E)
32. Ryan Wong, Necati Cihan Camgoz, and Richard Bowden. 2024. Sign2GPT: Leveraging large language models for gloss-free sign language translation. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Ryan+Wong%2C+Necati+Cihan+Camgoz%2C+and+Richard+Bowden%2E+2024%2E+Sign2GPT%3A+Leveraging+large+language+models+for+gloss%2Dfree+sign+language+translation%2E+In+International+Conference+on+Learning+Representations%2E)
33. Fuzhao Xue, Zian Zheng, Yao Fu, Jinjie Ni, Zangwei Zheng, Wangchunshu Zhou, and Yang You. 2024. [OpenMoE: An early effort on open mixture-of-experts language models](https://md.research.elamur.ai/free/en/papers/xue-2024-openmoe-early-effort-open-mixture/). In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pages…
34. Longhui Yuan, Binhui Xie, and Shuang Li. 2023. Robust test-time adaptation in dynamic scenarios. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 15922–15932. [find](https://scholar.google.com/scholar?q=Longhui+Yuan%2C+Binhui+Xie%2C+and+Shuang+Li%2E+2023%2E+Robust+test%2Dtime+adaptation+in+dynamic+scenarios%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+Computer+Vision+and+Pattern+Recognition+%28CVPR%29%2C+pages+1592)
35. Biao Zhang, Mathias Müller, and Rico Sennrich. 2023. SLTUNET: A simple unified model for sign language translation. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Biao+Zhang%2C+Mathias+M%C3%BCller%2C+and+Rico+Sennrich%2E+2023%2E+SLTUNET%3A+A+simple+unified+model+for+sign+language+translation%2E+In+International+Conference+on+Learning+Representations%2E)
36. Qingyang Zhang, Yatao Bian, Xinke Kong, Peilin Zhao, and Changqing Zhang. 2025. COME: Test-time adaption by conservatively minimizing entropy. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Qingyang+Zhang%2C+Yatao+Bian%2C+Xinke+Kong%2C+Peilin+Zhao%2C+and+Changqing+Zhang%2E+2025%2E+COME%3A+Test%2Dtime+adaption+by+conservatively+minimizing+entropy%2E+In+International+Conference+on+Learning+Representations%2E)
37. Benjia Zhou, Zhigang Chen, Albert Clapés, Jun Wan, Yanyan Liang, Sergio Escalera, Zhen Lei, and Du Zhang. 2023. Gloss-free sign language translation: Improving from visual-language pretraining. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 20871– 20881. [find](https://scholar.google.com/scholar?q=Benjia+Zhou%2C+Zhigang+Chen%2C+Albert+Clap%C3%A9s%2C+Jun+Wan%2C+Yanyan+Liang%2C+Sergio+Escalera%2C+Zhen+Lei%2C+and+Du+Zhang%2E+2023%2E+Gloss%2Dfree+sign+language+translation%3A+Improving+from+visual%2Dlanguage+pretraining%2E+In+Pro)
38. Hao Zhou, Wengang Zhou, Weizhen Qi, Junfu Pu, and Houqiang Li. 2021a. Improving sign language translation with monolingual data by sign back-translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 1316–1325. [find](https://scholar.google.com/scholar?q=Hao+Zhou%2C+Wengang+Zhou%2C+Weizhen+Qi%2C+Junfu+Pu%2C+and+Houqiang+Li%2E+2021a%2E+Improving+sign+language+translation+with+monolingual+data+by+sign+back%2Dtranslation%2E+In+Proceedings+of+the+IEEE%2FCVF+Conference+on+C)
39. Hao Zhou, Wengang Zhou, Yun Zhou, and Houqiang Li. 2021b. Spatial-temporal multi-cue network for sign language recognition and translation. IEEE Transactions on Multimedia, 24:768–779. [find](https://scholar.google.com/scholar?q=Hao+Zhou%2C+Wengang+Zhou%2C+Yun+Zhou%2C+and+Houqiang+Li%2E+2021b%2E+Spatial%2Dtemporal+multi%2Dcue+network+for+sign+language+recognition+and+translation%2E+IEEE+Transactions+on+Multimedia%2C+24%3A768%E2%80%93779%2E)
40. Xingzhi Zhou, Boyang Zhang, Zhiliang Tian, Yibo Zhang, Xin Niu, Ka Chun Cheung, Simon See, and Nevin L. Zhang. 2025. Resilient test-time adaptation by mitigating batch-normalization overfitting. In ICASSP 2025 – 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP),… [find](https://scholar.google.com/scholar?q=Xingzhi+Zhou%2C+Boyang+Zhang%2C+Zhiliang+Tian%2C+Yibo+Zhang%2C+Xin+Niu%2C+Ka+Chun+Cheung%2C+Simon+See%2C+and+Nevin+L%2E+Zhang%2E+2025%2E+Resilient+test%2Dtime+adaptation+by+mitigating+batch%2Dnormalization+overfitting%2E+In+IC)
41. Chenyang Zhu, Bin Xiao, Lin Shi, Shoukun Xu, and Xu Zheng. 2024. Customize Segment Anything Model for multi-modal semantic segmentation with mixture of LoRA experts. arXiv preprint arXiv:2412.04220. — [arXiv:2412.04220](https://arxiv.org/abs/2412.04220) [find](https://scholar.google.com/scholar?q=Chenyang+Zhu%2C+Bin+Xiao%2C+Lin+Shi%2C+Shoukun+Xu%2C+and+Xu+Zheng%2E+2024%2E+Customize+Segment+Anything+Model+for+multi%2Dmodal+semantic+segmentation+with+mixture+of+LoRA+experts%2E+arXiv+preprint+arXiv%3A2412%2E04220%2E+%E2%80%94)

*Updated 2 November 2026*