GeoAI, Deepfake Remote Sensing, and the Future of Earth Observation @sounny
GeoAI, Deepfake Remote Sensing, and the Future of Earth Observation  @sounny
Uploaded October 2026 | Updated November 2026, 2 days ago
What happens when you combine planetary satellite constellations with modern vision AI and autonomous coding agents?

In this lecture explores the shift from classical remote sensing and pixel classification (Random Forest, Maximum Likelihood) to Vision Language Models, local GeoAI agents, and the emerging threat of deepfake satellite imagery.

Timestamps / Chapters:
00:00 - Introduction & The Earth Observation Data Explosion
05:12 - Classical Remote Sensing: Landsat & Spectral Signatures
14:30 - Beyond the Visible Spectrum: Infrared, SWIR & Vegetation Indices
24:15 - Image Classification: Random Forest vs. Maximum Likelihood
34:40 - The Rebranding Trap: What ESRI and Google Call "GeoAI"
44:10 - Vision Models vs. Classical Math: The "Trees in the Parking Lot" Problem
53:25 - Spatial Awareness in LLMs: Emerging Geolocation Capabilities
1:02:15 - Building Local GeoAI Agents: Function Calling & Autonomous Map Making
1:08:40 - Deepfake Remote Sensing: From Colin Powell's UN Speech to Synthetic Geography
1:12:10 - The Future of the Spatial Analyst

Key Resources & Links:
• Personal Academic Hub: sounny.com
• Planetary Web GIS (JSMARS): jsmars.sounny.com
• Interactive GIS Web Tools: sounny.github.io
• LinkedIn: linkedin.com/in/sounny

#GeoAI #RemoteSensing #GIS #EarthObservation #SatelliteImagery #DeepLearning #ArtificialIntelligence #SpatialDataScience #Landsat #AgenticAI
GeoAI, Deepfake Remote Sensing, and the Future of Earth ObservationGIS Analysis on John Snow Map Part 2 - Proportional SymbolsGoogle Maps replaces Mercator Projection for a Globe!Geographic Weighted Regression on 911 phone callsLab 10 - Basics of Kriging and GeoStatsMapping US Census Tracts with web-GISMoroccan US Freetrade Agreement announced by George W. BushGIS on John Snow Map Part 3 – Standard Ellipse and Mean CenterTGV Maroc - ONCFMapping Datum Shifts from NAD27 to NAD83 Part 1From JSON to Silicon  Architecting Mission Critical Space SoftwSenseFly drones generate an amazing 3D model of the Matterhorn
Moulay Anwar Sounny-Slitine |

GeoAI, Deepfake Remote Sensing, and the Future of Earth Observation

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER