# Parameter-Efficient Adaptation of Geospatial Foundation Models through Embedding Deflection

arXiv: https://arxiv.org/abs/2503.09493 · HTML: https://mlsanity.com/paper/2503.09493/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 40 of its 62 references that are in the map.
- First posted on arXiv 12 Mar 2025.

## Builds on (most cited first, 30 of 40)

- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models](https://mlsanity.com/paper/2106.10199/) (2021)
- [Parameter-Efficient Fine-Tuning for Large Models: A Comprehensive Survey](https://mlsanity.com/paper/2403.14608/) (2024)
- [Compacter: Efficient Low-Rank Hypercomplex Adapter Layers](https://mlsanity.com/paper/2106.04647/) (2021)
- [SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery](https://mlsanity.com/paper/2207.08051/) (2022)
- [LoRA+: Efficient Low Rank Adaptation of Large Models](https://mlsanity.com/paper/2402.12354/) (2024)
- [Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning](https://mlsanity.com/paper/2212.14532/) (2022)
- [Training Neural Networks with Fixed Sparse Masks](https://mlsanity.com/paper/2111.09839/) (2021)
- [Rethinking Positional Encoding in Language Pre-training](https://mlsanity.com/paper/2006.15595/) (2020)
- [Composable Sparse Fine-Tuning for Cross-Lingual Transfer](https://mlsanity.com/paper/2110.07560/) (2021)
- [Panoptic Segmentation of Satellite Image Time Series with Convolutional Temporal Attention Networks](https://mlsanity.com/paper/2107.07933/) (2021)
- [xBD: A Dataset for Assessing Building Damage from Satellite Imagery](https://mlsanity.com/paper/1911.09296/) (2019)
- [Self-supervised Learning in Remote Sensing: A Review](https://mlsanity.com/paper/2206.13188/) (2022)
- [A Kernel-Based View of Language Model Fine-Tuning](https://mlsanity.com/paper/2210.05643/) (2022)
- [Towards Geospatial Foundation Models via Continual Pretraining](https://mlsanity.com/paper/2302.04476/) (2023)
- [An Empirical Study of Remote Sensing Pretraining](https://mlsanity.com/paper/2204.02825/) (2022)
- [Sparse Low-rank Adaptation of Pre-trained Language Models](https://mlsanity.com/paper/2311.11696/) (2023)
- [Self-Supervised Pretraining Improves Self-Supervised Pretraining](https://mlsanity.com/paper/2103.12718/) (2021)
- [Cross-Attention is All You Need: Adapting Pretrained Transformers for Machine Translation](https://mlsanity.com/paper/2104.08771/) (2021)
- [Satellite Image Time Series Classification with Pixel-Set Encoders and Temporal Self-Attention](https://mlsanity.com/paper/1911.07757/) (2019)
- [PANGAEA: A Global and Inclusive Benchmark for Geospatial Foundation Models](https://mlsanity.com/paper/2412.04204/) (2024)
- [Tuning LayerNorm in Attention: Towards Efficient Multi-Modal LLM Finetuning](https://mlsanity.com/paper/2312.11420/) (2023)
- [OmniSat: Self-Supervised Modality Fusion for Earth Observation](https://mlsanity.com/paper/2404.08351/) (2024)
- [A Billion-scale Foundation Model for Remote Sensing Images](https://mlsanity.com/paper/2304.05215/) (2023)
- [CMID: A Unified Self-Supervised Learning Framework for Remote Sensing Image Understanding](https://mlsanity.com/paper/2304.09670/) (2023)
- [A New Learning Paradigm for Foundation Model-based Remote Sensing Change Detection](https://mlsanity.com/paper/2312.01163/) (2023)
- [Cross-Scale MAE: A Tale of Multi-Scale Exploitation in Remote Sensing](https://mlsanity.com/paper/2401.15855/) (2024)
- [Self-supervised Vision Transformers for Joint SAR-optical Representation Learning](https://mlsanity.com/paper/2204.05381/) (2022)
- [One for All: Toward Unified Foundation Models for Earth Vision](https://mlsanity.com/paper/2401.07527/) (2024)

## Built on by (most cited first, 10 of 10)

- [Attention, Please! Revisiting Attentive Probing Through the Lens of Efficiency](https://mlsanity.com/paper/2506.10178/) (2025)
- [Fine-tune Smarter, Not Harder: Parameter-Efficient Fine-Tuning for Geospatial Foundation Models](https://mlsanity.com/paper/2504.17397/) (2025)
- [Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation](https://mlsanity.com/paper/2504.06962/) (2025)
- [Scaling Laws for Geospatial Foundation Models: A case study on PhilEO Bench](https://mlsanity.com/paper/2506.14765/) (2025)
- [SIMPLER: Efficient Foundation Model Adaptation via Similarity-Guided Layer Pruning for Earth Observation](https://mlsanity.com/paper/2603.19873/) (2026)
- [Landslide Hazard Mapping with Geospatial Foundation Models: Geographical Generalizability, Data Scarcity, and Band Adaptability](https://mlsanity.com/paper/2511.04474/) (2025)
- [CrossEarth-Gate: Fisher-Guided Adaptive Tuning Engine for Efficient Adaptation of Cross-Domain Remote Sensing Semantic Segmentation](https://mlsanity.com/paper/2511.20302/) (2025)
- [GeoRoPE: Ground-Aware Rotary Adaptation for Remote Sensing Foundation Models](https://mlsanity.com/paper/2606.14760/) (2026)
- [Multimodal Foundation Models Adaptation based on Domain-Aware Relaxed Orthogonal Subspace for Remote Sensing](https://mlsanity.com/paper/2609.13654/) (2026)
- [Super-Resolving Unseen Hyperspectral Sensors at Any Scale via Spatial Operators](https://mlsanity.com/paper/2609.39926/) (2026)

## Abstract (arXiv)

As large-scale heterogeneous data sets become increasingly available, adapting foundation models at low cost has become a key issue. Seminal works in natural language processing, e.g. Low-Rank Adaptation (LoRA), leverage the low "intrinsic rank" of parameter updates during adaptation. In this paper, we argue that incorporating stronger inductive biases in both data and models can enhance the adaptation of Geospatial Foundation Models (GFMs), pretrained on RGB satellite images, to other types of optical satellite data. Specifically, the pretrained parameters of GFMs serve as a strong prior for the spatial structure of multispectral images. For this reason, we introduce DEFLECT (Deflecting Embeddings for Finetuning Latent representations for Earth and Climate Tasks), a novel strategy for adapting GFMs to multispectral satellite imagery with very few additional parameters. DEFLECT improves the representation capabilities of the extracted features, particularly enhancing spectral information, which is essential for geoscience and environmental-related tasks. We demonstrate the effectiveness of our method across three different GFMs and five diverse datasets, ranging from forest monitoring to marine environment segmentation. Compared to competing methods, DEFLECT achieves on-par or higher accuracy with 5-10× fewer parameters for classification and segmentation tasks. The code will be made publicly available.