paper-with-me

Papers

Cross-sensor self-supervised training and alignment for remote sensing

2024-05-16 · Valerio Marsocci, Nicolas Audebert

Large-scale "foundation models" have gained traction as a way to leverage the vast amounts of unlabeled remote sensing data collected every day. However, due to the multiplicity of Earth Observation satellites, these models should learn "sensor agnostic" representations, that generalize across sensor characteristics with minimal fine-tuning. This is complicated by data availability, as low-resolution imagery, such as Sentinel-2 and Landsat-8 data, are available in large amounts, while very high-resolution aerial or satellite data is less common. To tackle these challenges, we introduce cross-sensor self-supervised training and alignment for remote sensing (X-STARS). We design a self-supervised training loss, the Multi-Sensor Alignment Dense loss (MSAD), to align representations across sensors, even with vastly different resolutions. Our X-STARS can be applied to train models from scratch, or to adapt large models pretrained on e.g low-resolution EO data to new high-resolution sensors, in a continual pretraining framework. We collect and release MSC-France, a new multi-sensor dataset, on which we train our X-STARS models, then evaluated on seven downstream classification and segmentation tasks. We demonstrate that X-STARS outperforms the state-of-the-art by a significant margin with less data across various conditions of data availability and resolutions.

📄 PDF Abstract BibTeX arXiv:2405.09922

Code (0)

등록된 구현이 없습니다.

Tasks

Continual PretrainingEarth Observation

Methods 이 논문이 사용한 방법론

ALIGN In the ALIGN method, visual and language representations are jointly trained from noisy image alt-text data. The image and text encoders are learned via contrastive loss…

Similar Papers 제목 키워드 기반

SEMI: Self-supervised Exploration via Multisensory Incongruity

2020-09-26 · Jianren Wang, Ziwen Zhuang, Hang Zhao

Efficient exploration is a long-standing problem in reinforcement learning since extrinsic rewards are usually sparse or missing. A popular solution to this issue is to feed an agent with novelty signals as intrinsic rew…

Efficient Exploration

How do Cross-View and Cross-Modal Alignment Affect Representations in Contrastive Learning?

2022-11-23 · Thomas M. Hehn, Julian F. P. Kooij, Dariu M. Gavrila

Various state-of-the-art self-supervised visual representation learning approaches take advantage of data from multiple sensors by aligning the feature representations across views and/or modalities. In this work, we inv…

Contrastive Learningcross-modal alignmentDepth EstimationDepth Prediction+5

Learning Transferable Sensor Models via Language-Informed Pretraining

2026-03-12 · Yuliang Chen, Arvind Pillai, Yu Yvonne Wu, Tess Z. Griffin 외 arxiv

Modern sensing systems generate large volumes of unlabeled multivariate time-series data. This abundance of unlabeled data makes self-supervised learning (SSL) a natural approach for learning transferable representations…

Self-Supervised LearningQuestion Answering

Robust Self-Supervised Cross-Modal Super-Resolution against Real-World Misaligned Observations

2026-02-21 · Xiaoyu Dong, Jiahuan Li, Ziteng Cui, Naoto Yokoya arxiv

Cross-modal super-resolution (SR) on real-world misaligned data is challenging, as only unlabeled low-resolution (LR) source and high-resolution (HR) guide images with complex spatial misalignment are available. Previous…

Unsupervised Domain Adaptation in Semantic Segmentation Based on Pixel Alignment and Self-Training

2021-09-29 · Hexin Dong, Fei Yu, Jie Zhao, Bin Dong 외

This paper proposes an unsupervised cross-modality domain adaptation approach based on pixel alignment and self-training. Pixel alignment transfers ceT1 scans to hrT2 modality, helping to reduce domain shift in the train…

Domain AdaptationSegmentationSemantic SegmentationUnsupervised Domain Adaptation