ZAYAN introduces feature-level zero-anchor contrastive pretraining that produces disentangled embeddings and improves classification accuracy on remote sensing tabular datasets over standard deep learning baselines.
arXiv:2204.09387 (2022)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ZAYAN: Disentangled Contrastive Transformer for Tabular Remote Sensing Data
ZAYAN introduces feature-level zero-anchor contrastive pretraining that produces disentangled embeddings and improves classification accuracy on remote sensing tabular datasets over standard deep learning baselines.