SpectralGPT: Spectral Remote Sensing Foundation Model
Authors:
- Danfeng HongDHDanfeng Hong
- Bing ZhangBing Zhang
- Xuyang LiXLXuyang Li
- Yuxuan LiYLYuxuan Li
- Chenyu LiCLChenyu Li
- Jing YaoJYJing Yao
- Naoto YokoyaNaoto Yokoya
- Hao LiHLHao Li
- Pedram GhamisiPGPedram Ghamisi
- Xiuping JiaXJXiuping Jia
- Antonio PlazaAPAntonio Plaza
- Paolo GambaPGPaolo Gamba
Preprint
Abstract
The foundation model has recently garnered significant attention due to its potential to revolutionize the field of visual representation learning in a self-supervised manner. While most foundation models are tailored to effectively process RGB images for various visual tasks, there is a noticeable gap in research focused on spectral data, which offers valuable information for scene understanding, especially in remote sensing (RS) applications. To fill this gap, we created for the first time a universal RS foundation model, named SpectralGPT, which is purpose-built to handle spectral RS images using a novel 3D generative pretrained transformer (GPT). Compared to existing foundation models, SpectralGPT 1) accommodates input images with varying sizes, resolutions, time series, and regions in a progressive training fashion, enabling full utilization of extensive RS big data; 2) leverages 3D token generation for spatial-spectral coupling; 3) captures spectrally sequential patterns via multi-target reconstruction; 4) trains on one million spectral RS images, yielding models with over 600 million parameters. Our evaluation highlights significant performance improvements with pretrained SpectralGPT models, signifying substantial potential in advancing spectral RS big data applications within the field of geoscience across four downstream tasks: single/multi-label scene classification, semantic segmentation, and change detection.Keywords
All Related Versions
This publication has 26 references indexed in Scilit:
- Cross-city matters: A multimodal remote sensing benchmark dataset for cross-city semantic segmentation using high-resolution domain adaptation networksRemote Sensing of Environment, 2023
- Masked Autoencoders Are Scalable Vision Learners2022
- Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote Sensing Data2021
- Exploring Simple Siamese Representation Learning2021
- Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised Visual Representation Learning2021
- Momentum Contrast for Unsupervised Visual Representation Learning2020
- Deep learning and process understanding for data-driven Earth system scienceNature, 2019
- Unified Perceptual Parsing for Scene Understanding2018
- Functional Map of the World2018
- Deep Residual Learning for Image Recognition2016