Scholarlink

This paper introduces SenPa-MAE, a transformer architecture that encodes the sensor parameters of an observed multispectral signal into the image embeddings. SenPa-MAE can be pre-trained on imagery of different satellites with non-matching spectral or geometrical sensor characteristics. To incorporate sensor parameters, we propose a versatile sensor parameter encoding module as well as a data augmentation strategy for the diversification of the pre-training dataset. This enables the model to effectively differentiate between various sensors and gain an understanding of sensor parameters and the correlation to the observed signal. Given the rising number of Earth observation satellite missions and the diversity in their sensor specifications, our approach paves the way towards a sensor-independent Earth observation foundation model. This opens up possibilities such as cross-sensor training and sensor-independent inference. The code for model training as well as the model weights are publicly available under https://github.com/JonathanPrexl/SenPa-MAE .

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SenPa-MAE: Sensor Parameter Aware Masked Autoencoder for Multi-satellite Self-supervised Pretraining

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jonathan Prexl,
  • Michael Schmitt

摘要

This paper introduces SenPa-MAE, a transformer architecture that encodes the sensor parameters of an observed multispectral signal into the image embeddings. SenPa-MAE can be pre-trained on imagery of different satellites with non-matching spectral or geometrical sensor characteristics. To incorporate sensor parameters, we propose a versatile sensor parameter encoding module as well as a data augmentation strategy for the diversification of the pre-training dataset. This enables the model to effectively differentiate between various sensors and gain an understanding of sensor parameters and the correlation to the observed signal. Given the rising number of Earth observation satellite missions and the diversity in their sensor specifications, our approach paves the way towards a sensor-independent Earth observation foundation model. This opens up possibilities such as cross-sensor training and sensor-independent inference. The code for model training as well as the model weights are publicly available under https://github.com/JonathanPrexl/SenPa-MAE .