ZAYAN: Disentangled Contrastive Transformer for Tabular Remote Sensing Data
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909004011143168 |
|---|---|
| author | Habib, Al Zadid Sultan Bin Tasnim, Tanpia Islam, Md. Ekramul Tabasum, Muntasir |
| author_facet | Habib, Al Zadid Sultan Bin Tasnim, Tanpia Islam, Md. Ekramul Tabasum, Muntasir |
| contents | Learning informative representations from tabular data in remote sensing and environmental science is challenging due to heterogeneity, scarce labels, and redundancy among features. We present ZAYAN (Zero-Anchor dYnamic feAture eNcoding), a self-supervised, feature-centric contrastive framework for tabular data. ZAYAN performs contrastive learning at the feature rather than sample level, removing the need for explicit anchor selection and any reliance on class labels, while encouraging a redundancy-minimized, disentangled embedding space. The framework has two modules: ZAYAN-CL, which pretrains feature embeddings via a zero-anchor contrastive objective with dynamic perturbations and masking, and ZAYAN-T, a Transformer that conditions on these embeddings for downstream classification. Across eight datasets, including six remote-sensing tabular benchmarks and two remote-sensing-driven flood-prediction tables from satellite and GIS products, ZAYAN achieves superior accuracy, robustness, and generalization over tabular deep learning baselines, with consistent gains under label scarcity and distribution shift. These results indicate that feature-level contrastive learning and dynamic feature encoding provide an effective recipe for learning from tabular sensing data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_27606 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ZAYAN: Disentangled Contrastive Transformer for Tabular Remote Sensing Data Habib, Al Zadid Sultan Bin Tasnim, Tanpia Islam, Md. Ekramul Tabasum, Muntasir Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition Learning informative representations from tabular data in remote sensing and environmental science is challenging due to heterogeneity, scarce labels, and redundancy among features. We present ZAYAN (Zero-Anchor dYnamic feAture eNcoding), a self-supervised, feature-centric contrastive framework for tabular data. ZAYAN performs contrastive learning at the feature rather than sample level, removing the need for explicit anchor selection and any reliance on class labels, while encouraging a redundancy-minimized, disentangled embedding space. The framework has two modules: ZAYAN-CL, which pretrains feature embeddings via a zero-anchor contrastive objective with dynamic perturbations and masking, and ZAYAN-T, a Transformer that conditions on these embeddings for downstream classification. Across eight datasets, including six remote-sensing tabular benchmarks and two remote-sensing-driven flood-prediction tables from satellite and GIS products, ZAYAN achieves superior accuracy, robustness, and generalization over tabular deep learning baselines, with consistent gains under label scarcity and distribution shift. These results indicate that feature-level contrastive learning and dynamic feature encoding provide an effective recipe for learning from tabular sensing data. |
| title | ZAYAN: Disentangled Contrastive Transformer for Tabular Remote Sensing Data |
| topic | Machine Learning Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.27606 |