UTICA: Multi-Objective Self-Distllation Foundation Model Pretraining for Time Series Classification
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_ | 1866908859247886336 |
|---|---|
| author | Moakher, Yessin Hili, Youssef Attia El Feofanov, Vasilii |
| author_facet | Moakher, Yessin Hili, Youssef Attia El Feofanov, Vasilii |
| contents | Self-supervised foundation models have achieved remarkable success across domains, including time series. However, the potential of non-contrastive methods, a paradigm that has driven significant advances in computer vision, remains underexplored for time series. In this work, we adapt DINOv2-style self-distillation to pretrain a time series foundation model, building on the Mantis tokenizer and transformer encoder architecture as our backbone. Through a student-teacher framework, our method Utica learns representations that capture both temporal invariance via augmented crops and fine-grained local structure via patch masking. Our approach achieves state-of-the-art classification performance on both UCR and UEA benchmarks. These results suggest that non-contrastive methods are a promising and complementary pretraining strategy for time series foundation models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_01348 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | UTICA: Multi-Objective Self-Distllation Foundation Model Pretraining for Time Series Classification Moakher, Yessin Hili, Youssef Attia El Feofanov, Vasilii Machine Learning Artificial Intelligence Self-supervised foundation models have achieved remarkable success across domains, including time series. However, the potential of non-contrastive methods, a paradigm that has driven significant advances in computer vision, remains underexplored for time series. In this work, we adapt DINOv2-style self-distillation to pretrain a time series foundation model, building on the Mantis tokenizer and transformer encoder architecture as our backbone. Through a student-teacher framework, our method Utica learns representations that capture both temporal invariance via augmented crops and fine-grained local structure via patch masking. Our approach achieves state-of-the-art classification performance on both UCR and UEA benchmarks. These results suggest that non-contrastive methods are a promising and complementary pretraining strategy for time series foundation models. |
| title | UTICA: Multi-Objective Self-Distllation Foundation Model Pretraining for Time Series Classification |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2603.01348 |