Expanding sample sizes, diversifying sample types, enhancing computational capabilities

The gap

Evidence profile

Sourced from the inline gaps and conclusions of the source papers, classified as general, spanning 2 journals.

Supporting evidence — 3 representative gaps

• MFK-Net: a computationally efficient Mamba-Fourier-KAN hybrid architecture for UAV-based crop classification (2026) · Frontiers in Plant Science · doi

  While cross-dataset experiments on PlantDoc and Martell Forest demonstrate generalization to distinct domains, the model’s performance on crops from other geographic regions (such as temperate cereal crops or Mediterranean orchard species) remains to be validated. Such integration could enable recognition of novel crop species not present in the training data, addressing a key limitation of super- vised approaches.

  generalinline gaps
  Keywords: crops species cross dataset experiments plantdoc martell forest demonstrate generalization distinct domains model performance geographic
• Parcel-level crop classification in small and irregular fields using structured Sentinel-2 data and clustering based analysis (2026) · Scientific Reports · doi

  While this study focused on four regions and a selected set of crop types, and the spatial and temporal transferability of the framework was not yet explored, these aspects point to valuable directions for further validation across additional regions, crops, and years toward broader application. While the masking-and-interpolation augmentation improved the recall of minority classes and thus the overall Macro F1, it was insufficient to resolve the confusion among the most spectrally overlapping summer crops, as it reproduces existing temporal patterns rather than introducing new spectral diversity.

  generalconclusions
  Keywords: regions temporal crops focused four selected crop types spatial transferability framework explored aspects point valuable
• Tea tree recognition based on multi-source satellite data across Southeast China (2026) · Frontiers in Plant Science · doi

  Future studies should focus on expanding sample sizes, diversify- ing sample types, enhancing computational capabilities, optimiz- ing model architecture, and implementing multi-faceted, multi- tiered validation methods to achieve more precise and detailed investigations into crop recognition issues in both southern and northern regions.

  generalinline gaps
  Keywords: sample multi future focus expanding sizes diversify types enhancing computational capabilities optimiz model architecture implementing

Questions about this gap