PhenoYieldNet: Multi-Crop Yield Prediction via Crop Phenology
Researchers propose PhenoYieldNet, a framework for multi-crop yield prediction that models crop-specific phenological responses to weather. It uses a Crop Phenology Bank and Crop Phenology Attention module to capture temporal patterns. The method aims to improve generalization across diverse crop types, addressing limitations of single-crop models. The paper is available on arXiv.
Key facts
- PhenoYieldNet is a multi-crop yield prediction framework
- It models crop-specific phenology with temporal drivers
- Includes a Crop Phenology Bank (CPB) and Crop Phenology Attention (CPA) module
- CPB integrates learnable embeddings guided by queries
- CPA captures multi-scale trend and variation components
- Aims to improve generalization across diverse crop types
- Addresses limitations of single-crop prediction models
- Paper available on arXiv with ID 2605.23478
Entities
Institutions
- arXiv