Computer vision models predict land surface displacement for carbon capture

Researchers have developed CarbonNet, a novel approach utilizing computer vision to predict land surface displacement from subsurface geometry images for carbon capture and storage (CCS) projects. This method aims to reduce the high computational costs and generalization limitations associated with traditional physics-based models. The study implemented and compared various models, including CNN, ResNet, ResNetUNet, LSTM, and Transformer++, finding that ResNetUNet performed best for static mechanics problems and LSTM showed comparable results to Transformer++ for transient scenarios. AI

IMPACT This research demonstrates a novel application of computer vision for optimizing carbon capture and storage, potentially accelerating climate change mitigation efforts.

RANK_REASON Research paper published on arXiv detailing a new application of computer vision for climate change mitigation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• carbon capture and storage
• Carbonnet
• climate change
• CNN
• computer vision
• LSTM
• ResNetUNet
• Transformer++
• Wei Chen

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →