CR-JEPA: Cross-Modal Joint-Embedding Predictive Learning for Remote Sensing Image Retrieval
Not on the map. Why?
Abstract (arXiv)
Cross-modal remote sensing image retrieval aims to retrieve semantically related scenes across heterogeneous sensing modalities. This remains challenging because paired observations may differ substantially in imaging physics, spatial resolution, spectral configuration, and visual appearance. Moreover, a single retrieval projection trained with one objective may be insufficient to jointly support cross-modal semantic alignment and same-modal neighbourhood preservation. We propose CR-JEPA, a Cross-modal Retrieval Joint-Embedding Predictive Architecture for dual-modality remote sensing retrieval. The model uses modality-specific stems, a shared transformer trunk, and JEPA-style predictive objectives to estimate masked latent target features within and across modalities. Inspired by LeJEPA, we apply Sketched Isotropic Gaussian Regularization to raw retrieval projections to stabilize embeddings and mitigate collapse. CR-JEPA further employs a decoupled-head design with a unified retrieval head for same-modal retrieval and a cross-modal retrieval head for cross-modal search. We evaluate CR-JEPA on BEN-14K, CBRSIR_VS, and DSRSID. On BEN-14K, CR-JEPA improves S1 to S2 retrieval from 61.23% to 75.82% and S2 to S1 retrieval from 63.73% to 75.40% over X-JEPA, while also achieving competitive same-modal retrieval with fewer parameters.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 22 of its 42 references that are in the map. First posted on arXiv 30 May 2026.
Builds on
22 of 42 references.
References · 22
- 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
- 2021VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learningcited by 1,057
- 2023Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecturecited by 1,039
- 2022SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagerycited by 408
- 2022Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learningcited by 308
- 2025LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristicscited by 254
- 2019BigEarthNet: A Large-Scale Benchmark Archive For Remote Sensing Image Understandingcited by 234
- 2023SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagerycited by 205
- 2023CROMA: Remote Sensing Representations with Contrastive Radar-Optical Masked Autoencoderscited by 169
- 2021BigEarthNet-MM: A Large Scale Multi-Modal Multi-Label Benchmark Archive for Remote Sensing Image Classification and Retrievalcited by 92
- 2022SSL4EO-S12: A Large-Scale Multi-Modal, Multi-Temporal Dataset for Self-Supervised Learning in Earth Observationcited by 84
- 2024Rethinking Transformers Pre-training for Multi-Spectral Satellite Imagerycited by 83
10 more
- 2024AnySat: One Earth Observation Model for Many Resolutions, Scales, and Modalitiescited by 74
- 2024Cross-Scale MAE: A Tale of Multi-Scale Exploitation in Remote Sensingcited by 47
- 2023Decoupling Common and Unique Representations for Multimodal Self-supervised Learningcited by 36
- 2024Exploring Masked Autoencoders for Sensor-Agnostic Image Retrieval in Remote Sensingcited by 12
- 2022A Novel Self-Supervised Cross-Modal Image Retrieval Method In Remote Sensingcited by 10
- 2019CMIR-NET : A Deep Learning Based Model For Cross-Modal Retrieval In Remote Sensingcited by 7
- 2025REJEPA: A Novel Joint-Embedding Predictive Architecture for Efficient Remote Sensing Image Retrievalcited by 4
- 2025CSMoE: An Efficient Remote Sensing Foundation Model with Soft Mixture-of-Expertscited by 3
- 2026GeoMeld: Toward Semantically Grounded Foundation Models for Remote Sensingcited by 2
- 2026From Alignment to Prediction: A Study of Self-Supervised Learning and Predictive Representation Learningcited by 2
Built on by
Not cited yet.