# JEDI: JEPA-to-Edge Distillation for Efficient Cropland Segmentation from Satellite Imagery

arXiv: https://arxiv.org/abs/2609.07915 · HTML: https://mlsanity.com/paper/2609.07915/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Detection and dense prediction distillation, one of 12 directions in Knowledge distillation.
- Builds on 22 of its 29 references that are in the map.
- First posted on arXiv 7 Sep 2026.

## Placed in

- Knowledge distillation / [Detection and dense prediction distillation](https://mlsanity.com/distill/detection-and-dense-prediction-distillation/)

## Builds on (most cited first, 22 of 22)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [Masked-attention Mask Transformer for Universal Image Segmentation](https://mlsanity.com/paper/2112.01527/) (2021)
- [Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture](https://mlsanity.com/paper/2301.08243/) (2023)
- [Contrastive Representation Distillation](https://mlsanity.com/paper/1910.10699/) (2019)
- [SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery](https://mlsanity.com/paper/2207.08051/) (2022)
- [Structured Knowledge Distillation for Dense Prediction](https://mlsanity.com/paper/1903.04197/) (2019)
- [Training Very Deep Networks](https://mlsanity.com/paper/1507.06228/) (2015)
- [Channel-wise Knowledge Distillation for Dense Prediction](https://mlsanity.com/paper/2011.13256/) (2020)
- [Foundation Models for Generalist Geospatial Artificial Intelligence](https://mlsanity.com/paper/2310.18660/) (2023)
- [Cross-Image Relational Knowledge Distillation for Semantic Segmentation](https://mlsanity.com/paper/2204.06986/) (2022)
- [Foreground-Aware Relation Network for Geospatial Object Segmentation in High Spatial Resolution Remote Sensing Imagery](https://mlsanity.com/paper/2011.09766/) (2020)
- [A Good Student is Cooperative and Reliable: CNN-Transformer Collaborative Learning for Semantic Segmentation](https://mlsanity.com/paper/2307.12574/) (2023)
- [CalCROP21: A Georeferenced multi-spectral dataset of Satellite Imagery and Crop Labels](https://mlsanity.com/paper/2107.12499/) (2021)
- [Can SAM recognize crops? Quantifying the zero-shot performance of a semantic segmentation foundation model on generating crop-type maps using satellite imagery for precision agriculture](https://mlsanity.com/paper/2311.15138/) (2023)

## Abstract (arXiv)

Large vision models provide useful representations for remote-sensing segmentation but are often too expensive for deployment at the satellite or field edge. Existing feature-level distillation methods also tend to assume similar teacher and student architectures and often stop feature alignment when task training begins. We introduce JEDI (JEPA-to-Edge Distillation), a two-stage framework that transfers representations from a large I-JEPA Vision Transformer teacher to a compact SegFormer student. First, JEDI aligns the student's terminal representation with the teacher's token space using cross-architecture projection and spatial alignment. It then jointly optimizes supervised segmentation, temperature-scaled response distillation, and persistent feature alignment throughout task adaptation. On CalCROP21, JEDI-B0 achieves 68.0 mean Intersection-over-Union (mIoU) with 4.04M parameters, improving over the standalone student by 16.0 points and coming within 2.0 points of the 70.0 mIoU achieved by the 639M-parameter teacher. We evaluate SegFormer B0, B1, and B2 students with 4.04M, 14.33M, and 28M parameters, respectively. Across all three variants, JEDI consistently outperforms response-, structure-, channel-, and relational-distillation baselines under the same teacher-student setting. These results show that persistent representation alignment is especially valuable under aggressive compression, substantially reducing model size and computation while preserving segmentation performance.