AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

SpectralGPT: Spectral Remote Sensing Foundation Model

Danfeng Hong, Bing Zhang, Xuyang Li, Yuxuan Li, Chenyu Li, Jing Tao Yao, Naoto Yokoya, Hao Li, Pedram Ghamisi, Xiuping Jia, Antonio Plaza, Paolo E. Gamba and 2 more

TPAMI 2024 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 1,117 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add SpectralGPT pretraining on top of random init (patch 8) (ablation)ViT-Base · BigEarthNetmAP: 80.15 -> 85.61Improved
use patch size 8 instead of 16 (ablation)SpectralGPT · BigEarthNetmAP: 76.42 -> 85.61Improved
pretrain on fMoW-S2 instead of BigEarthNet only (ablation)SpectralGPT · BigEarthNetmAP: 83.97 -> 85.61Improved
add BigEarthNet continuation pretraining after fMoW-S2 (ablation)SpectralGPT · BigEarthNetmAP: 85.61 -> 87.51Improved
raise masking ratio from 25% to 90% (ablation)SpectralGPT · BigEarthNetmAP: 82.73 -> 85.61Improved
raise masking ratio from 75% to 90% (ablation)SpectralGPT · BigEarthNetmAP: 84.64 -> 85.61Improved
normalize the reconstruction target (ablation)SpectralGPT · BigEarthNetmAP: 84.95 -> 85.61Improved
normalization instead of standardization (ablation)SpectralGPT · BigEarthNetmAP: 85.39 -> 85.61Improved
decoder depth 4 instead of 2 (ablation)SpectralGPT · BigEarthNetmAP: 84.59 -> 85.61Improved
increase decoder depth from 4 to 8 blocks (ablation)SpectralGPT · BigEarthNetmAP: 85.61 -> 85.38Hurt
scale ViT-Base to ViT-HugeSpectralGPT · BigEarthNetmicro-mAP: 85.61 -> 89.23Improved
add BigEarthNet to pretraining dataSpectralGPT · SegMunichmIoU: 49.8 -> 51.0Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy