RAIC Labs (formerly Synthetaic)’s Post

Our new article from the 2023 AAG GeoAI Panel is now published online: https://lnkd.in/evXsNcYu We discussed the advances and limitations of GeoAI research in the past five years, and potential directions for the next five years and beyond. #AAG #GeoAI #geospatial

Today, we kick off the 15th European Conference on Synthetic Aperture Radar (EUSAR), in my eyes the most inportant conference on SAR. Besides the tutorial on machine learning for SAR I am going to present today, my team is actively contributing with several talks and posters. Be there or be square! #sar #eusar #syntheticapertureradar #remotesensing #earthobservation #radar #deeplearning Thomas Roßberg is discussing multi-frequency SAR and its relation to vegetation indices. Francescopaolo Sica is talking about self-supervised learning for phase and coherence estimation in InSAR. Philipp Sibler is presenting is latest work on deep learning-based InSAR data simulation. Michael Recla is showing the latest progress in single-image height estimation and fusion of ascending/descending orbit data. And I will demonstrate what single-image height estimation is good for besides just DEM generation.

There is an old saying that if want to go fast, go alone; if you want to go far, go together. However, when two lean and agile companies like SpecTIR and MetaSpectral join forces, just maybe we can go farther and faster. The amount of data generated by an HSI system is voluminous. To date SpecTIR has handled the data volume and velocity the old fashion way – hours and hours of manual processing and analysis. However as new HSI air and space systems come online, new automated methods will need to be developed and employed to keep up with the coming Hellabytes of data. After 25+ years of advancing hyperspectral science, and perfecting the art, SpecTIR knows a thing or two about the power of hyperspectral imaging and where we need to be. By integrating MetaSpectral's AI Fusion platform, we aim to go farther and faster and push the boundaries of hyperspectral imaging. Join us in exploring a VSWIR data cube of the Yerington area in Nevada in our joint Spectral Sandbox: https://lnkd.in/gY8VRE6S #Hyperspectral #AIforAnalytics #FutureOfData #Metaspectral #SpecTIR  #hyperspectralimaging #HSI #remotesensing #geology, #LWIR #SWIR #thermalInfrared #spectiraustralia #exploration, #minerals #environmentalmonitoring #explorationgeology

The paper titled BMSU-Net : Building Extraction From Arial Imagry received the "Best Paper Award" at the 5th International Conference on Innovative Trends In Information Technology (ICITIIT) organized by the Indian Institute of Information Technology (IIIT Kottayam), under the IEEE. The paper presents our novel deep learning model designed for predicting buildings from aerial images. This innovation of our research in the domain of computer vision and its applications in urban planning and infrastructure development. The Project is also under Indian Institute of Remote Sensing (IIRS) from our institute. #SJCE #IIITKottayam #ICITIIT #IEEE #IIRS #ISRO #DeepLearning

: constant flow : Continuation of installation style explorations. #comfyui #aianimation #genai #installation #aicc #stablediffusion Stability AI #animatediff

GeoAI continues to be a hot topic at #EsriFedGIS2024. The session presenters were quick to point out that there are more GeoAI tools in #ArcGIS than you think. I know I’ve used pretrained models for feature extraction from imagery. Check out the many pretrained models available that you can use now. #geoai #Esri #spatialanalysis #imagery https://lnkd.in/eYQ99fyn

📢 Don't miss our final presentation of the day, starting SOON! ✨ We'll dive deep into geologic formation tops prediction and how to seamlessly integrate machine learning workflows into your subsurface analysis. Presentation Times: Sept 23 & 24: 3:10 PM Sept 25: 1:10 PM ATCE 2024 | #GeologicFormation #ML #DataVisualization #IVAAP #ATCE2024

🌍 Want to travel the world with Intuition-1 satellite? No problem! In the coming weeks, we will share images from different locations, showcasing the capabilities of our Leopard Data Processing Unit (DPU) in action. 🌿 Today, we present a capture of the Culuene River in Brazil, highlighting diverse vegetation along its banks. Different shades of red and green in the image represent vegetation intensity, which is crucial for environmental monitoring and resource management. 📸 The image uses hyperspectral imaging, capturing light across 192 spectral bands from visible to near-infrared light, unlike traditional imaging, which uses only three color bands. CIR (Color Infrared) combines infrared, red, and green bands to show healthy vegetation in shades of red, helpful in assessing plant health and monitoring crops. 🚀 Launched in November 2023, Intuition-1 is equipped with the Leopard DPU, which processes hyperspectral data in orbit using AI, performing up to three trillion operations per second. This enables real-time image segmentation and classification, reducing the volume of data transmitted to Earth and enhancing efficiency. 🌊 The Culuene River supports local agriculture and biodiversity. Its meandering course, formed by erosion and sediment deposition, is captured in detail by Intuition-1. This mission aims to prove that using AI to process hyperspectral data in orbit improves remote sensing efficiency by detecting phenomena undetectable by standard satellites and prioritizing key data for transmission. #KPLabs #dataprocessingunit #data #satelite #missioncomplete #CIR #Intuition1

I am thankful to #MadisonMagazine (I think) for this very nice piece showing the power of #OSINT, the #SPARTAIV case, and the work done with Royal United Services Institute and Geollect! Most importantly, the piece remarks my appreciation for #geometry, #ImageryAnalysis and #algebra. At the end of the day, people, yeah, IA is cool, but the human brain is better - and yes, excellent tools out there, but nothing will replace the power of the main disciplines. #GEOINT, #OSINT

Completed the MOOC of Going Places with Spatial Analysis to learn about arcgisonline and its useful tools. #esri #esriMOOC #spatialanalysis #arcgisonline