Asterisk Labs is expanding its team of climate and Earth observation scientists. You could help us build an alternative to the usual routes for doing research, prioritising positive societal impact, and democratic worker control!. (1/4)

🛰️ 🔭 Say hello to Galileo, an open-source multimodal model designed to process many kinds of earth observations at once—driving applications such as mapping agricultural land, detecting floods, and monitoring marine pollution.

My first model was a model that had a human in the loop to help craft responses to real-time chats. I link to that lecture in this post (this was 2017 🤯).

One major insight that will be lost in the hype of AlphaEarth is MOSAIKS (Rolf et al. 2021) is a top performer in the paper's experiments while requiring ZERO pretraining - yes you read that right. No petabyte-scale pretraining dataset, no cluster of A100/H100s. Just

Our new AI model AlphaEarth Foundations is mapping the planet in astonishing detail. 🌏🔍. Scientists will now be able to track the impact of deforestation, monitoring crop health, and more – significantly faster, thanks to our new datasets. 🧵

Super happy our work, “EcoMapper: Generative Modeling for Climate-Aware Satellite Imagery” will be at #ICML2025! EcoMapper generates satellite views under different climate scenarios by utilising millions of Sentinel-2 images 🛰️🚀

🚀 New Feature Alert: Instance Segmentation with GeoAI! 🛰️🧠. The GeoAI Python package now supports instance segmentation for remote sensing imagery — all with just a few lines of code!. You can quickly:.✅ Train instance segmentation models.✅ Run inference on multiple

🚀 GeoAI Tutorial 24: Water Detection Using GeoAI with Sentinel-2 Satellite Imagery 🌍🛰️ . In this tutorial, I walk you through training a model to detect water bodies from Sentinel-2 satellite imagery with 6 spectral bands — then show you how to run batch inference across

New @MistralAI Cookbook: 🛰️ Finetuning Pixtral on a satellite imagery dataset 🛰️. - How to call Mistral's batch inference API.- How to pass images (encoded in base64) in your API calls to Mistral's VLM (here Pixtral-12B).- How to fine-tune Pixtral-12B on an image classification

🔥Achieved 91.74 IoU!. 🚀Existing methods struggle to fully tap into low-level features in #RemoteSensing images—limiting building extraction accuracy. Enter CFF-Net: a Cross-stage Features Fusion Network that changes the game. 📖Read more: #ComputerVision

Attractive!!! Over 2156 views on a groundbreaking solution in #RemoteSensing! Current ML/DL training samples for imagery lack an integrated approach. Enter #LuoJiaSET, a large - scale database with over five million samples, solving key challenges. Link: