A platform meant for demystifying Spatial data, GIS & Remote sensing as well as associated technologies such as Artificial intelligence, Machine Learning & how they interlink.

- how much area was affected by a disaster. - what's the area affected by deforestation.- how much harvest did a farm produce etc. A geo foundation model is trained on large earth observation datasets available to answer this and such questions.

AI Foundation Models for Geospatial and Earth observation: A New Era of Earth Understanding. In the geospatial space, the questions are earth related e.g. - where is mining activity.- how many buildings are in an area.

Heard of Foundation models? I'll assume yes given that LLMs are a type foundation models. What of GeoFoundation Models?. Foundation Models are large scale AI models trained on massive datasets to answer a myriad of questions.

AI agents are powered by LLMs & leverages an LLM to manage workflow execution & make decisions, recognizes when a workflow is complete & can proactively correct its actions if needed. When it fails, it can halt execution & transfer control back to the user.

AI Agents are systems that independently accomplish tasks on your behalf ; automate workflows on the users behalf where a workflow is a sequence of steps that must be executed to meet the user's goal e.g. resolving a customer issue, booking a reservation. reference:OpenAI guide

Pandas in Machine Learning & Deep learning.

Libraries for working with data in Machine Learning & Deep learning: Pandas- Part 1.

Pandas is a python library for data wrangling(converting data into a useful formal) and analysis. It allows one to load/read various types of data. Find a blog post and a colab notebook inspired by the image on Pandas operations. ⬇️.

Google Deepmind project Astra is an AI model that takes in the world. First mentioned at Google I/O 2024 as a universal agent useful in everyday life, It uses natural interaction & has a multimodal understanding of the world around us through audio & vision. A demo at 60minutes

The model's repo as shared by @OurRadiantEarth .

RAMP architecture is based off Eff-Unet combining the effectiveness of compound scaled EfficientNet as the encoder for feature extraction with UNet decoder for reconstructing the fine-grained segmentation map.

More on Replicable AI for Microplanning (RAMP), a segmentation model for extracting buildings. RAMP model card provides details on:.- The model overview- its description and methodology.- Considerations towards the best performance.- Model Performance expressed in IOU & F1 score

RT @robmarkcole: GeoDeep is a Python package that can detect objects in satellite imagery. It's made up of 1,026 lines of Python and uses O….

Bellwether: first prediction engine for the Earth & everything on it.Focused on:.- wildfire risk modeling & mitigation tools.- insurance for natural disasters.- scene assessment tools to support first responders after natural disasters.

WeatherNext, a family of AI models from Google DeepMind and Google Research that produces state-of-the-art weather forecasts.

**Towards a Trajectory-powered Foundation Model of Mobility** that advocates for a geospatial foundation model based on human mobility trajectories in the built environment i.e. transportation networks, data-driven urban planning, tourism & sustainability.

Then there's the population Dynamics foundation model(PDFM) for insights into population dynamics: A foundation model for geospatial inference.

The newest model-Geospatial Reasoning for unlocking insights with generative AI and multiple foundation models. It aims to provide comprehensive answers about our world, complete with detailed plans and data visualizations:.

Google has several geospatial foundation models as shared in their Geospatial Reasoning Interest Form. For anyone curious about the links to each of the models, check under this thread

RT @GoogleAI: We just announced Geospatial Reasoning, a new research effort that will use generative AI (#Gemini) to help unlock novel insi….