🚀 Ryan Bank introduces Vexcel Intelligence! Great AI needs great data. But until now, AI simply couldn't see Earth at the high-resolution level where real insight happens. The Vexcel Data Program changes that! You can now search Planet Earth for features, detect patterns, understand context and track change over time. 🗺️ Vexcel delivers highly accurate, high-resolution aerial imagery and geospatial products across 45+ countries — offering more detail than satellites and broader reach than drones. 👉 Curious about Vexcel Intelligence? Learn more: https://bit.ly/41rgs1e #AI #Geospatial #ComputerVision #EarthObservation #MachineLearning #VexcelImaging #AerialImagery #UltraCam #innovation
More Relevant Posts
-
In the latest episode of our podcast, we dived into the technology behind forest monitoring 🌲 Earth observation satellites, drones, ground-based monitoring networks: together, they are able to provide quite large datasets on forest conditions. As explained by Alba Viana-Soto, Earth Observation scientist at the Technical University of Munich, the challenge lies in properly processing and integrating these datasets through the use of AI and computational algorithms capable of handling such large volumes of data. Listen to the episode on ICONS Innovation Strategies Spotify channel 🎧 https://lnkd.in/d2H8Tr-2 #TheSoundOfForests
To view or add a comment, sign in
-
it’s amazing the amount of performance that can be gained right now by trading-in a better dense feature extractor in all sorts of DNN architectures
Today my colleagues at Land & Carbon Lab and AI at Meta and I are releasing a huge update to our high-resolution global canopy height map (Tolan et al. 2024). The updated dataset introduces three major improvements: • DINOv2 → DINOv3 backbone (+0.11 R²) • Expanded and improved training data (+0.12 R²) • Optimized training parameters (+0.10 R²) When we have high-quality input imagery (leaf-on, sun elevation >45°, off-nadir <25°), the model can begin to resolve horizontal canopy structure, closely matching airborne LiDAR canopy height models, such as the example below from Kalimantan, Indonesia. Read the preprint here: https://lnkd.in/ebQpWVHp Access the data here: https://lnkd.in/eW9fPicG Download the model weights / code here: https://lnkd.in/e-HTPYmg
To view or add a comment, sign in
-
-
Paper Title: Machine Learning and Deep Temporal Networks for UAV Velocity Classification using High-Dimensional CSI and RSSI Data Authors: Ankit Kumar 1, Sacharitha Sirisilla 1, Abhay Saxena 2 and Yadangi Abhishek 2, 1University of Agder, Norway, 2Indian Institute of Technology Bhubaneswar, India Abstract With increasing utilization of Unmanned Aerial Vehicles in civilian and industrial scenarios, reliable velocity estimation is imperative for safe tracking and airspace security. This work presents a 5G-centric RF sensing framework that jointly leverages Received Signal Strength Indicator and Channel State Information using a USRP-based SDR testbed under controlled line-of-sight flights of drones. More than 19 million RSSI samples and high-dimensional CSI measurements were collected over multiple velocities. For RSSI-based inference, the proposed custom 1D-CNN+BiGRU model significantly outperforms classical machine learning baselines. For CSI-based inference, CNN-LSTM and ensemble models-like Random Forest and XGBoost-powerfully capture velocity-dependent channel variations. Results show that RSSI provides coarse temporal cues while CSI encodes fine spatial-multipath structure, and hybrid deep temporal models generalize best. The proposed dual-modality framework is scalable and easily extensible to UAV detection, localization, and intent-aware surveillance. Keywords Channel State Information (CSI), Real-time classification, Wireless Signal Analysis, Time-Series Data, Temporal dependencies, RSSI, Drone velocity identification, Experimental Setup, 5G, Deep learning Volume URL: https://lnkd.in/dvsN5ggR Abstract URL: https://lnkd.in/gyDfhWMP Pdf URL: https://lnkd.in/gUSBkCi4 #ChannelStateInformation #Realtimeclassification #WirelessSignalAnalysis #TimeSeriesData #Temporaldependencies #RSSI #Dronevelocityidentification #ExperimentalSetup #5G #Deeplearning #callforpapers #researchpapers #cfp #researchers #phdstudent #artificialintelligence #deeplearning #machinelearning #generativeai #bigdata #databases #researchscholar #journalpaper #submission #journalsubmission
To view or add a comment, sign in
-
-
New Post: ## Enhanced Drone Swarm Coordination via Decentralized Predictive Resource Allocation for Remote Emergency Medical Delivery - **Abstract:** Delivering emergency medical supplies to remote mountainous regions using autonomous drones faces challenges related to weather unpredictability, terrain complexity, and dynamic demand. This paper proposes a novel decentralized predictive resource allocation \(DPRA\) system leveraging graph neural networks \(GNNs\) and reinforcement learning \(RL\) to optimize drone swarm coordination. DPRA dynamically predicts resource needs, assigns drones \[…\]
To view or add a comment, sign in
-
War enters the model. Modern militaries now rely on agentic AI running chain-of-thought reasoning to interpret signals from drones, satellites, radar, and city sensors. Signals in. Patterns out. Decisions in milliseconds. But once agentic AI enters the mission loop, a new attack surface emerges. Adversaries no longer need to destroy infrastructure. They only need to deceive the model’s reasoning. A poisoned GPS signal. A manipulated sensor feed. A prompt hidden inside an image. And in the next phase of conflict, the most dangerous weapon may be the lie the AI learns to trust. My latest piece: “The First AI Battlefield: War Enters the Model” (Link in comments) #AgenticAI #DefenseTech #AISecurity #NationalSecurity #ArtificialIntelligence
To view or add a comment, sign in
-
-
New Dataset Drop for Precision Agriculture + Vision AI! BaleUAVision introduces a high-quality UAV dataset for hay bale detection and counting—a surprisingly underexplored yet critical task for yield estimation and logistics. Built from 2,599 high-res RGB images across 16 real fields, it captures diverse flight conditions (altitude, speed, overlap) and includes dense polygon annotations in COCO, YOLO, and segmentation formats. Why it matters: • Real-world variability → better generalization • Supports detection and instance segmentation • Designed to complement foundation models like SAM • Strong benchmark results (YOLOv11 shows high precision/recall across regions & altitudes) If you’re working on UAV vision, edge AI, or agri-tech—this is a dataset worth exploring! Congrats to the team! Georgios Karatzinis Socratis Gkelios Athanasios Kapoutsis ------------------------------------------------------------------ In our Vision AI weekly newsletter, we cover the latest updates in the Vision AI space. Interested to know more? Link below 👇 #ComputerVision #VisionAI #AI #ML #Agtech #YOLO
To view or add a comment, sign in
-
-
🌍 #Maps are no longer just maps. They are becoming #intelligent. With the rise of Retrieval-Augmented Generation (#RAG), geospatial data is entering a new era of AI-powered location intelligence. Every day, massive volumes of spatial data are generated through #satellites, #drones, #GISsystems, and #sensors. The challenge is not just collecting this data it is making sense of it quickly and accurately. This is where #RAG meets #Geospatial. By retrieving relevant spatial datasets and combining them with AI models, systems can now generate contextual insights for real-world decision making. Imagine asking: ➡️ Which areas are most vulnerable to flooding? ➡️ Where should a new smart city infrastructure be developed? ➡️ Which agricultural zones need immediate intervention? And receiving data-backed answers powered by satellite imagery, terrain models, and #GIS layers. From #disastermanagement and #climateresilience to #smartcities, #agriculture, #logistics, and #spacetechnologies, the integration of #AI + #Geospatial is redefining how we understand our #planet. At NeoGeoInfo Technologies Limited, we are excited about the possibilities of leveraging RAG-driven geospatial intelligence to turn complex spatial data into actionable insights. The future of mapping is not just visual. It is intelligent, contextual, and AI-driven. #GeoAI #RAG #Geospatial #LocationIntelligence #AIInnovation #SatelliteData #SmartCities #NeoGeoInfo
To view or add a comment, sign in
-
-
New Post: Advanced Multispectral UAV Sensor Fusion for Early‑Stage Nitrogen Deficiency Detection in Corn Using Convolutional Neural Networks and Kalman Filtering - https://lnkd.in/g4fAjatQ Multispectral UAV Sensor Fusion for Early‑Stage Nitrogen Deficiency Detection in Corn Using Convolutional Neural Networks and Kalman Filtering — ### Abstract Early detection of nitrogen \(N\) deficiency in corn fields is critical for optimizing fertilizer use and maximizing yield. We propose a sensor‑fusion framework that combines high‑resolution multispectral UAV imagery with ground‑truth soil assays \[…\]
To view or add a comment, sign in
-
New Post: Field‑Level Understory Carbon Stock Estimation via Integrated LiDAR‑NIR Remote Sensing and Deep Learning - https://lnkd.in/gFhwswhR ### Abstract Accurate quantification of understory vegetation carbon in temperate mixed‑species forests is a persistent bottleneck in forest‑biomass carbon accounting. Conventional plot‑based inventories are labor‑intensive, spatially sparse, and limited in aerial perspective, yielding high uncertainties in coarse‑scale assessments. We propose a novel, fully automated framework that fuses airborne Light Detection and Ranging \(LiDAR\) data \[…\]
To view or add a comment, sign in
-
The future of infrastructure intelligence is becoming spatial. With the rise of GIS, AI models, and Generative AI, we are moving from static maps to dynamic decision systems. Data from drones, satellites, and sensors can now be analyzed to detect patterns, predict risks, and support faster planning. In sectors like energy and infrastructure, this shift is turning geospatial data into real operational insight. Excited to see how AI-powered geospatial intelligence will redefine how we monitor, plan, and manage large-scale assets. #GIS #GenerativeAI #SpatialData #GeoAI #EnergyInfrastructure
To view or add a comment, sign in