SPECTRA: Band-Routed Embedding and Stage-Wise LoRA for Cross-Sensor Fine-Tuning of Geospatial Foundation Models
This work addresses the practical challenges of spectral mismatch and high fine-tuning costs for users adapting pre-trained geospatial foundation models to new Earth observation datasets, offering an incremental improvement in efficiency and performance.
This paper tackles the problem of adapting geospatial foundation models (GeoFMs) to new sensor data, which often have different spectral bands than the pre-trained models expect, while also reducing the computational cost of fine-tuning. The authors propose SPECTRA, a framework that uses Band-Routed Embedding (BRE) to map all available downstream bands into the expected band space and Stage-wise Transferability-aware LoRA (ST-LoRA) to efficiently fine-tune by allocating LoRA ranks based on transferability. Experiments across three GeoFMs and four segmentation datasets demonstrate that BRE improves performance by utilizing all spectral bands, and ST-LoRA reduces trainable parameters compared to full fine-tuning and standard LoRA.
Geospatial foundation models (GeoFMs), pretrained on large-scale geospatial data such as Earth observation (EO), climate, and weather data, have shown promising performance when fine-tuned on diverse downstream tasks. However, there are two challenges of adapting EO-pretrained GeoFMs to practical downstream datasets. The first challenge is how to handle spectral mismatch: pretrained patch embeddings expect a fixed set of input bands, whereas downstream sensors may provide different channels. The second challenge is how to reduce fine-tuning cost and make it efficient. While existing work has made efforts on these challenges individually, jointly improving fine-tuning performance under spectral mismatch while reducing adaptation cost remains underexplored. We propose SPECTRA, a parameter-efficient fine-tuning framework that addresses both spectral mismatch and adaptation cost. To handle spectral mismatch, SPECTRA introduces Band-Routed Embedding (BRE), which maps all available downstream bands into the band space expected by the pretrained GeoFM. By using BRE, all available bands in the downstream dataset are utilized to improve the selected-band input without changing the pretrained patch embedding interface. To reduce adaptation cost, SPECTRA further introduces a Stage-wise Transferability-aware LoRA (ST-LoRA) fine-tuning. ST-LoRA estimates stage-wise transferability before fine-tuning and assigns stage-specific LoRA ranks, concentrating trainable parameters on the stages with high transferability for the target task. Across three EO-pretrained GeoFMs and four downstream segmentation datasets, experiments show that BRE improves performance by utilizing all spectral bands, while ST-LoRA reduces trainable parameters compared with full fine-tuning and standard LoRA. Code is available at https://github.com/big-data-lab-umbc/SPECTRA.